3.1 Composition and textural characteristics of MoO3 and MoO3/TiO2‒SiO2 samples
The XRD patterns of all three samples are shown in Fig. 1. According to the XRD results, samples 1 and 2 include α-MoO3 of orthorhombic modification. The orthorhombic phase α-MoO3 has a layered structure consisting of double layers of MoO6 octahedra with covalent bonds in planes (100) and (001), and van der Waals forces between layers in direction (010). In the double-layer structure, oxygen is bound to a single molybdenum sticking out from both sides of the double layer [30]. The interaction between the layers is much weaker than the interaction between the molybdenum and oxygen inside the layer. The unit cell parameters of α-MoO3 are а = 3.964 Å, b = 13.863 Å and с = 3.696 Å for sample 1 (MoO3/TiO2–SiO2 composites) and a = 3.961 Å, b = 13.869 Å, c = 3.699 Å for sample 2 (MoO3 sphere). The average crystallite size of samples 1 and 2 are 256 Å and 384 Å, respectively. The lack of possibility to detect diffraction maxima belonging to titanium dioxide of anatase structure on the diffractogram of MoO3/TiO2–SiO2 sample (Fig. 1, sample 1) may be due to both the small amount of this oxide and the fact that its most intense peak appears at the same angles as those of molybdenum trioxide. The presence of silicon oxide in the MoO3/TiO2–SiO2 composites does not confirm by XRD, that can be due to the small amount of SiO2.
The TiO2–SiO2 sample is TiO2 of anatase structure (Fig. 1, sample 3) with the parameters: a = b = 3.776 Å, c = 9.418 Å. Wide diffraction peaks in the X-ray patterns confirm the low degree of crystallinity of this sample, that probably deals with the presence of X-ray-amorphous silicon dioxide. The presence of SiO2 in the sample 3 is proved by the results of EDS analysis. As can be seen from the X-ray emission spectrum of TiO2–SiO2 sample (Fig. 2, a), silicon is present in addition to Ti and O.
EDS analysis of MoO3/TiO2–SiO2 sample (Fig. 2, b) also indicates the presence of Si and Ti in addition to Mo and O elements, that confirms the presence not only molybdenum oxide, but also titanium oxide and silicon oxide in this sample.
The results of SEM indicate that the samples had a spherical shape with the same size as that of the TOKEM-320 anion exchanger, which was used as the template (Fig. 3).
MoO3 and MoO3/TiO2‒SiO2 spheres consisted of rod-like structure agglomerates with the length from 1 to 7 µm, which indicates the aggregation of crystallites (data from the Scherrer‘s Eq. 256 Å and 384 Å). The morphology of the connected microrods indicates that the microcrystals of these samples had a hierarchical structure. However, significant morphological changes between samples depending on their composition were not observed. The results of 3D computer microtomography confirm that the MoO3/TiO2‒SiO2 spherical composites are hollow inside. Two structures can be visually distinguished in Fig. 4, the dark color refers to the air-filled cavity of the sphere, and the lighter color refers to the oxide skeleton of spherical composites.
The layering of the hierarchical spherical structure is proved by the results of qualitative X-ray microanalysis of the composite in section (Fig. 5). As can be seen from the distribution of Mo, O and Ti elements, the molybdenum trioxide constitutes the core of the sphere, and TiO2–SiO2 composite is the shell in the form of a thin layer. The element Si in the distribution profile was not observed due to its low initial amount in the sol.
The MoO3 and MoO3/TiO2‒SiO2 spheres can be attributed to mesoporous samples. As shown in Fig. 6a, the adsorption/desorption isotherm of the MoO3 samples show that they are mesoporous materials. The same pattern was observed for MoO3/TiO2‒SiO2 samples. The coating of the spherical MoO3 core by the TiO2‒SiO2 layer did not lead to changes in the type of sorption, porosity and specific surface area (5.8 m2/g) of the samples (Fig. 6).
IR spectra of the spherical MoO3 and MoO3/TiO2‒SiO2 samples are shown in Fig. 7. The absorption bands corresponding to water and oxygen molecules adsorbed from the air on the surface of the samples were not detected in the IR spectra.
The spectrum of MoO3 contains two intense and two weak peaks of valence vibrations of molybdenum‒oxygen bonds of the oxide crystal structure. Short [Mo = O] double bond appears as a band at 970 cm−1. IR absorptions at 812 and 825 cm−1 refer to the valence vibrations of the [Mo‒O‒Mo] bridge bonds. The assignment of the bond vibrational bands was performed in accordance with the literature data [31]. In the IR spectrum of the MoO3/TiO2‒SiO2 sample (Fig. 7, b) vibrations of bonds belonging to [Ti‒O] of the TiO2 matrix at 430 and 500 cm−1 are not observed [32]. The absorptions around 800, 960 cm−1, which refer to the [Si‒O‒Si] bond vibrations and [Si‒O‒Ti] heterobonding vibrations [33, 34], respectively, was not detected, since they overlap with the region of molybdenum‒oxygen bond vibrations [35, 36]. However, the shift of [Mo = O] vibration at 978 cm−1 as well as [Mo‒O‒Mo] vibrations at 825 and 855 cm−1 to a higher frequency region indirectly indicates the presence of these bonds.
The comparison of the IR spectra of the samples shows that the only one difference among them is the pronounced absorption maximum at 1099 cm−1 in the spectrum of MoO3/TiO2-SiO2 composites (Fig. 7, b). According to the literature data, these absorption bands can be attributed to the [Si‒O‒Mo] bond vibrations.
The Raman spectra of the samples are shown in Fig. 8. The Raman active modes were identified in accordance to the literature data [37] and confirmed the orthorhombic phase of MoO3. The similarity of the Raman spectra of MoO3 and MoO3/TiO2‒SiO2 samples indicates that the crystal structure of molybdenum trioxide is preserved during the formation of TiO2‒SiO2 layer on it. In the spectrum of MoO3/TiO2-SiO2 composite in addition to the MoO3 modes, the most intense Raman peaks of the TiO2 structure of anatase at 114.0, 126.8, and 153.9 cm−1 are present. Such a number of modes can be explained by the fact that the TiO2 consists of particles of different sizes and the shift of the frequencies of these modes to the long-wavelength region indirectly indicates the formation of a bond Mo‒O‒Ti.
Full width change at half maximum (FWHM) of the Raman peak at 819 cm−1 in the MoO3/TiO2-SiO2 sample compared to the FWHM of the Raman peak at 819 cm−1 in the MoO3 sample (Fig. 8) indicates that the defectiveness of the MoO3 structure in these samples is different. As can be seen from Fig. 8, FWHM decreases from MoO3 sample to MoO3/TiO2‒SiO2 sample. This indicates a greater number of oxygen vacancies in the MoO3 than in the MoO3/TiO2-SiO2 sample. The intensity ratio of Raman modes at 282‒290 cm−1 and the equation (x = 0.01296R + 2.93429) obtained from Dieterle's data [38] were used to determine the oxygen stoichiometry in MoO3 − x. Table 1 shows that the MoO3 sample has more vacancies compared to the MoO3/TiO2‒SiO2 sample. This result is consistent with the FWHM of the Raman peak at 819 cm−1.
Table 1
Values of the ratio of intensity of the Raman peaks at 282–290 cm−1 with their respective oxygen vacancies.
Sample
|
R(I283/ I290)
|
R(I282/ I289)
|
Oxygen index x (MoOx)
|
Oxygen vacancy
|
MoO3
|
1.6538
|
-
|
2.9556
|
0.0444
|
MoO3/TiO2-SiO2
|
-
|
1.4166
|
2.9525
|
0.0425
|
Consequently, MoO3/TiO2‒SiO2 sample (sample 1) is a composite of α-MoO3, anatase phase of TiO2 and amorphous phase of SiO2 and sample 2 is α-MoO3. The formation of hierarchical layered spherical MoO3/TiO2‒SiO2 composites of shell/core type by obtained by template and sol-gel methods at 500 °C using TOKEM-320Y occurs through the formation of interphase boundaries between orthorhombic phase of MoO3 and TiO2‒SiO2 composite which consists of TiO2 of tetragonal structure and amorphous phase of SiO2, and formation of Ti‒O‒Mo and Si‒O‒Mo chemical bonds. The use of the sol based on tetrabutoxytitanium with tetraethoxysilane contributes to obtaining the regular (dense) TiO2‒SiO2 coating on the spherical hollow core of MoO3. The TiO2‒SiO2 layer reduces the number of oxygen vacancies in the MoO3 structure.
3.2 Photocatalytic properties of MoO3 and MoO3/TiO2-SiO2 samples
UV–visible spectroscopy was used to study the light absorption properties of samples (Fig. 9A). The experimental results of absorptions measurements of MoO3, TiO2–SiO2 and MoO3/TiO2–SiO2 suspensions were analyzed by the function (α⋅h⋅ν)2 = A2(h⋅ν − ΔEg) [39]. Experimental results of measurements of the dependence α from h⋅ν for indirect inter band transitions are represented as plots of √α versus h⋅ν (Fig. 9B), and a scheme for determining the optical bandgap energy (ΔEg) of MoO3, TiO2–SiO2 and MoO3/TiO2–SiO2 samples was also plotted.
Figure 9 shows that the MoO3 sample demonstrates stronger absorption intensity than the MoO3/TiO2–SiO2 sample in the hard ultraviolet region. The absorption of MoO3 sample decreases if it is covered by the TiO2–SiO2 layer, that is probably the result of the fundamental process of the formation of an electron–hole pair upon irradiation [40]. The MoO3/TiO2–SiO2 composite absorbs more light than the MoO3 sample in the soft ultraviolet region. The determined ΔEg values are 3.18, 2.33 and 2.58 eV for MoO3, TiO2–SiO2 and MoO3/TiO2–SiO2 samples, respectively. These values indicate that the formation of TiO2–SiO2 layer on MoO3 core leads to a decrease in the bandgap energy.
The photocatalytic activity of MoO3 and MoO3/TiO2–SiO2 was investigated in the degradation of methylene blue (MB) under light irradiation. The results of photocatalytic measurements are presented in Fig. 10. As can be seen in Fig. 10A, the color and absorption intensity of the pure dye change little with increasing irradiation time. The degree of degradation of pure MB reaches 14 wt. % in 120 min of irradiation with wave length at 312 nm. Keeping the MB with photocatalysts in the dark leads to the sorption of the dye on the surface of samples. The MoO3 composite adsorbed 76 wt. % of MB during 60 min while the MoO3/TiO2–SiO2 composite adsorbed 82 wt.%. It is can be explained by the fact that at equal pore diameters the volume of pores in the MoO3/TiO2–SiO2 composite is larger than in the MoO3 sample (Fig. 6). The change in the absorption spectrum of MS solution (Fig. 10B, C) indicates that the dye can undergo degradation in the presence of both samples.
As can be seen in Fig. 11, the photocatalytic decomposition of MB occurred for 40 min in the presence of MoO3 sample and for 20 min in the presence of MoO3/TiO2–SiO2 composite. The linear relationship between −ln(C/C0) and irradiation time can be describe by pseudo-first-order kinetics (−ln(C/C0) = −kt). The good linearity of the curves confirmed the applicability of this mode with a correlation coefficient value of 0.9621–0.9687 as shown in Fig. 11. Calculated values of the rate constants from the slope of the plot indicate that the rate of decomposition of MB in the presence of MoO3/TiO2–SiO2 (k = 0.0645 min−1) is about 2 times higher than that in the presence of MoO3 (k = 0.0336 min−1).
Chromatograms of MB solution after holding it in the dark in the presence of the catalyst (Fig. 12, a) showed no presence of peaks at lower retention times. This result indicates that only sorption of MB on the catalyst surface occurs in the dark. Chromatograms of MB solution in the presence of catalysts after UV light irradiation with wave length at 312 nm indicate that the discoloration of the solution is associated with a change in the structure of the organic dye (Fig. 12, b, c). This is evident by the decrease in the MB peak intensity at retention times of 12.12 min and 17.54 min and the appearance of new peaks detected at lower retention times, which correspond to new photocatalytic products of the dye. The comparison of the chromatographic separation spectra of MB solutions in the presence of different catalysts after irradiation showed (Fig. 12b, c) that MB practically does not remain in present of MoO3/TiO2‒SiO2 samples. This result confirms the results of study the photocatalytic activity of samples by spectrophotometry.
Further analysis was done by mass spectroscopic (MS) studies of the degradation intermediate products of MB solution after irradiation in present of MoO3/TiO2‒SiO2 sample and the results are shown in Fig. 13. In mass spectrum of the major detected degradation intermediates eluting at 12.8 min peak with m/z = 283.9 was observed which corresponds to the molar mass of the initial MB. It was not possible to separate the MB degradation intermediates eluting at 2.7 min, because they gather in one peak.
However, by mass spectrometry it was possible to record mass fragments with m/z = 270.8, 256.4 and 241.5, which indicate the presence of intermediate products of MB decomposition:
The formation of azure A, B and C and thionin through the demethylation cleavage during the photocatalytic degradation of MB agree with those reported in the literature [41].
Possible mechanism of photocatalytic actions can be assumed by determining the band structures of the components of the MoO3/TiO2–SiO2 composite. The valance band (VB) and conduction band (CB) potentials of the semiconductor can be calculated using the following empirical equations [42]:
$${E}_{VB}=X+\text{0,5}{E}_{g}-{E}_{e}$$
;
$${E}_{CB}={E}_{VB}-{E}_{g}$$
;
$$X={\left(X\left({A)}^{a}X\right({B)}^{b}\right)}^{\frac{1}{a+b}}$$
,
where EVB is the VB edge potential, X is the electronegativity of the semiconductor (a and b are the atomic number of compounds), Ee is the energy of free electrons vs hydrogen (4.5 eV), Eg is the band gap energy of the semiconductor, and ECB is the CB edge potential.
The results of calculating of the band gap, valence and conductivity band potentials of MoO3/TiO2-SiO2 are presented in Fig. 14.
The CB potential of MoO3 is ‒0.89 eV and more negative than that of TiO2–SiO2 (‒0.46 eV). Hence there is diffusion of electrons from the CB of MoO3 to TiO2–SiO2. At the same time, there is transfer of holes from the VB of TiO2–SiO2 (2.79 eV) to MoO3 (2.29 eV) because the valance band potential of the former is more positive than that of the latter. Thus the photogenerated electrons and holes are efficiently separated at the heterojunctions. The separated electrons released and the resulting holes are involved in the oxidation process of MB.