Nano TiO2 and Molybdenum/Tungsten Iodide Octahedral Clusters: Synergism in UV/Visible-Light Driven Degradation of Organic Pollutants

Emissions of various organic pollutants in the environment becomes a more and more acute problem in the modern world as they can lead to an ecological disaster in foreseeable future. The current situation forces scientists to develop numerous methods for the treatment of polluted water. Among these methods, advanced photocatalytic oxidation is a promising approach for removing organic pollutants from wastewater. In this work, one of the most common photocatalysts—titanium dioxide—was obtained by direct aqueous hydrolysis of titanium (IV) isopropoxide and impregnated with aqueous solutions of octahedral cluster complexes [{M6I8}(DMSO)6](NO3)4 (M = Mo, W) to overcome visible light absorption issues and increase overall photocatalytic activity. XRPD analysis showed that the titania is formed as anatase-brookite mixed-phase nanoparticles and cluster impregnation does not affect the morphology of the particles. Complex deposition resulted in the expansion of the absorption up to ~500 nm and in the appearance of an additional cluster-related band gap value of 1.8 eV. Both types of materials showed high activity in the photocatalytic decomposition of RhB under UV- and sunlight irradiation with effective rate constants 4–5 times higher than those of pure TiO2. The stability of the catalysts is preserved for up to 5 cycles of photodegradation. Scavengers’ experiments revealed high impact of all of the active species in photocatalytic process indicating the formation of an S-scheme heterojunction photocatalyst.


Introduction
Currently, one of the most discussed problems in the world is the depletion of water resources and pollution of soils, water, and air [1,2]. These problems are associated with rapid industrialization and the growth of production capacities. The majority of plants consume enormous amounts of water and transform it into wastewater, which contains various inorganic and organic species. Approximately 80% of this wastewater is released into the environment without any prior treatment [3]. Among hazardous pollutants, organic dyes account for the largest share [4]. These substances are widely used in various industries, such as tanning, textile, paper, and printing industries, and can be considered one of the most harmful for the environment [5,6]. Besides high stability and such direct effects as carcinogenicity and mutagenicity [7], dyes are dangerous in long-term consequences due to high extinction. Higher absorption of colored water diminishes the amount of light in depth affecting the photosynthesis process and, henceforth, decreasing the overall amount of dissolved oxygen [8]. Modern methods for removal of such pollutants, such as coagulation, precipitation, and adsorption, have several disadvantages forcing researchers to search for alternative methods [9]. The major disadvantages of coagulation and precipitation are the where D is crystallites size in nm, K is Scherrer constant and equal to 0.9, λ is the wavelength of the X-ray source in nm (Cu Kα = 0.15406 nm), β is full width at half maximum of the peak in radians, Θ is peak position in radians. Metal content was determined using inductively coupled plasma atomic emission spectroscopy (ICP-AES) on a ThermoScientific spectrometer iCAP-6500. The relative error is about 5-10%. Diffuse reflectance spectra were recorded using a UV-Vis-NIR 3101 PC spectrophotometer (Shimadzu Corporation, Kyoto, Japan). The emission spectra (λ ex = 320 nm) were recorded for powdered materials placed between two non-fluorescent glass plates with Agilent Cary Eclipse Fluorescence Spectrophotometer.
X-ray photoelectron spectroscopy (XPS) and valence band XPS (VB XPS) was performed on X-ray photoelectron spectrometer SPECS (SPECS Surface Nano Analysis GmbH, Berlin, Germany) using nonmonochromatic Al Kα irradiation (hν = 1486.6 eV). A thin layer of the powdered sample was applied onto conductive double-sided copper tape (3M ™ , Electron Microscopy Sciences, Hatfield, PA, USA). The binding energy scale was preliminarily calibrated by the position of the photoelectron lines of the core levels of gold (Au4f 7/2 , 84.0 eV), silver (Ag3d 5/2 , 368.3 eV), and copper (Cu2p 3/2 , 932.7 eV). To calibrate the recorded spectra, the Ti2p 3/2 line (BE = 458.8 eV) from the TiO 2 matrix was used as an internal standard.
An analysis of the porous structure was performed by a nitrogen adsorption technique using Quantochrome's Autosorb iQ at 77K. Initially, the materials were activated in a dynamic vacuum at 200 • C for 2 h. N 2 adsorption-desorption isotherms were measured within the range of relative pressures of 10 −6 to 0.995. The specific surface area was calculated from the data obtained on the basis of the conventional BET and DFT models. Pore size distributions were calculated using the DFT method.

Synthesis of Pure TiO 2
The titanium dioxide was obtained by the hydrolysis of titanium (IV) isopropoxide (Ti( i OPr) 4 ) [28][29][30]. The titanium (IV) isopropoxide was added dropwise to the hot distilled water (80 • C) at vigorous stirring in a 1:50 volume ratio. The resulting suspension was immediately sonicated for 5 min and then kept stirred at 80 • C for 4 h. Resulting white powder was washed 3 times with water and 2 times with acetone. The pure TiO 2 was dried in air at room temperature. Yield:~100%.

Synthesis of n x @TiO 2 (n = 1 (Mo) or 2 (W))
Cluster-containing materials n x @TiO 2 (where n = 1 (Mo) or 2 (W); x is the mass of the complex in gram per 1 g of TiO 2 and equal to 0.1, 0.5, 1, and 1.5) were obtained by impregnation of TiO 2 with the solution of cluster complex [{M 6 I 8 }(DMSO) 6 ](NO 3 ) 4 (M = Mo or W) in water. In a typical experiment, the required amount of the corresponding cluster was dissolved in 100 mL of water and 1 g of pure TiO 2 was added to the solution at constant vigorous stirring. The resulting suspensions were stirred for 20 h. Light yellow Nanomaterials 2022, 12, 4282 4 of 16 powders of n x @TiO 2 were washed several times with water until the washings became colorless and then 3 times with acetone. Materials were dried in air at room temperature.

Photocatalytic Experiment
Photocatalytic activity of n x @TiO 2 materials was investigated in the photodegradation of Rhodamine B (RhB) under UV irradiation (λ = 365 ± 5 nm,~13 mW cm −2 ). In a typical photocatalytic test, 20 mg of n x @TiO 2 and 60 mL of water were mixed in a quartz reactor and sonicated for 5 min. Then 20 mL of RhB solution (C = 10 mg L −1 ) was added to the reactor and the resulting mixture was stirred for 2 h in the dark to reach adsorptiondesorption equilibrium. The overall volume is 80 mL, RhB concentration is 2.5 mg L −1 , n x @TiO 2 concentration is 0.25 g L −1 . After that, the reaction mixture was irradiated with UV light for 45-120 min depending on the photocatalytic activity of the sample. During the irradiation, several aliquots of solutions (8 mL) were collected and centrifuged. UV-vis spectra of the isolated solutions were recorded in order to estimate the concentration of RhB. The decrease in the RhB concentration was tracked by its characteristic absorbance at 554 nm. The reactions rate constants (k eff ) were determined as pseudo-first order kinetics by linear approximation of ln(C/C 0 ) vs. t plot where C is the concentration of RhB at corresponding t, C 0 is the initial concentration of RhB, and t is the time when aliquots of solutions were taken. The effect of catalyst concentration was studied using 10 or 30 mg of n 0.1 @TiO 2 (n 0.1 @TiO 2 concentrations are 0.125 or 0.375 g L −1 ).

Cyclic Experiments
Cyclic experiments were performed in a similar manner as a typical photocatalytic test. After each run of irradiation (60 min) the aliquot (8 mL) was taken, centrifuged, and the UV-vis spectrum of the isolated solution was recorded. The remaining precipitate was placed back into the reaction mixture and the initial concentration of RhB (2.5 mg L −1 ) and overall volume (80 mL) were adjusted by the addition of an RhB solution.

Photodegradation of RhB under Solar Light
To study photodegradation under sun irradiation, a typical photocatalytic experiment was performed on a clear day in the middle of august (sunlight power =~30-35 mW cm −2 ). The reaction mixture in the quartz reactor without stirring was exposed to solar irradiation for 90-120 min depending on the photocatalytic activity of the sample. The air temperature was~25 • C.

Scavengers
Photocatalytic activity of n 0.1 @TiO 2 was assessed using the same procedure as earlier but in presence of scavengers for active species: Na 2 C 2 O 4 (h + , C = 10 mM), AgNO 3 (e − , C = 10 mM), i-PrOH (OH • , C = 10 mM). For the determination of O 2 •− impact the reaction mixture was preliminarily deaerated by bubbling with Ar gas for 20 min. Relative activity (RA) was calculated according to the following equation: where k eff (scav) is an effective rate constant in the presence of a certain scavenger and k eff (NS) is an effective rate constant in a scavenger free experiment.

Results and Discussion
3.1. Synthesis of Pure TiO 2 and n x @TiO 2 (n = 1 (Mo) or 2 (W)) Titanium dioxide was obtained by a simple hydrolysis of titanium (IV) isopropoxide in a neutral aqueous medium, i.e., by addition of Ti( i OPr) 4 into great excess of water (volume ratio is 1:50). For additional homogenization the resulting dispersion was sonicated immediately after the addition of titanium precursor. To transform partially amorphous TiO 2 into a crystalline phase the dispersion was heated at 80 • C for 4 h. According to XRPD analysis ( Figure 1G), the resulting TiO 2 powder is predominantly anatase phase with the admixture of brookite. The broadening of the peaks in the diffractogram indicates the formation of small particles. Using Scherrer Equation (1) we calculated the average size of TiO 2 particles, which is equal to 4.3 ± 1.0 nm. High dilution of the reaction mixture and low temperature probably results in the formation of extremely small crystalline mixed-phase particles. Nanomaterials 2022, 12, x FOR PEER REVIEW 5 of 16 (volume ratio is 1:50). For additional homogenization the resulting dispersion was sonicated immediately after the addition of titanium precursor. To transform partially amorphous TiO2 into a crystalline phase the dispersion was heated at 80 °C for 4 h. According to XRPD analysis ( Figure 1G), the resulting TiO2 powder is predominantly anatase phase with the admixture of brookite. The broadening of the peaks in the diffractogram indicates the formation of small particles. Using Scherrer Equation (1) we calculated the average size of TiO2 particles, which is equal to 4.3 ± 1.0 nm. High dilution of the reaction mixture and low temperature probably results in the formation of extremely small crystalline mixed-phase particles. Cluster-containing materials n x @TiO2 (n = 1 (Mo) or 2 (W); x is the mass of the complex in gram per 1 g of TiO2 and equal to 0.1, 0.5, 1, and 1.  Cluster-containing materials n x @TiO 2 (n = 1 (Mo) or 2 (W); x is the mass of the complex in gram per 1 g of TiO 2 and equal to 0.1, 0.5, 1, and 1.5) were obtained by impregnation of TiO 2 dispersion with a certain amount of cluster complexes [{Mo 6 4 (2) in water. XRPD analysis of the resulting yellowish powders demonstrates an absence of cluster-related peaks, indicating deposition of complex in amorphous state regardless of composition of cluster core ( Figures 1G and S2). Based on our experience and reaction conditions we believe that cluster is deposited on the TiO 2 surface in the form of fully hydrolyzed neutral aqua-hydroxo complexes [{M 6 I 8 }(H 2 O) 2 (OH) 4 ]·nH 2 O. The absence of any ligand-related peaks in FTIR spectra also confirms the full hydrolysis of the clusters ( Figure S3). The particle size of n x @TiO 2 calculated from diffractogram is close and almost independent of the amount of complex nor the type of cluster core, but slightly higher, than that of native TiO 2 , 4.3 ± 1.0 nm (TiO 2 ) vs. 4.8 ± 0.7 (1 0.1 @TiO 2 ) and 4.7 ± 0.8 (2 0.1 @TiO 2 ) nm.

Morphology and Composition
The morphology of the initial TiO 2 and n x @TiO 2 was studied using TEM ( Figure 1A-C).
One can see that all of the samples consist of small, uniformly distributed particles forming agglomerates of~100 nm or higher in diameter, indicating no influence of cluster impregnation on the morphology of the TiO 2 . Statistical analysis of the particle size ( Figure 1D-F) results in normal distribution curves with maxima of 4.8 ± 0.9 nm for TiO 2 , 5.0 ± 0.8 nm for 1 0.1 @TiO 2 , and 7.7 ± 1.3 nm for 2 0.1 @TiO 2 , which is close to the data from XRPD. The content of the molybdenum or tungsten in n x @TiO 2 was determined using ICP-AES and then converted to the content of {M 6 I 8 } per 1 g of TiO 2 ( Figure 1H, Table S1). According to the data obtained, the molybdenum cluster impregnates TiO 2 far less than the tungsten one. Even at x = 0.1, the content of {W 6 I 8 } is in order of magnitude higher than the content of {Mo 6 I 8 }, which is in agreement with the larger particle size of 2 0.1 @TiO 2 , determined from TEM. The subsequent increase in molybdenum cluster complex loading (x) does not lead to the increase in the {Mo 6 I 8 } content in 1 x @TiO 2 . On the contrary, the increase in the loading of the tungsten complex results in a linear increase in the {W 6 I 8 } content in the 2 x @TiO 2 materials. Such difference in the interaction of the clusters with TiO 2 is most likely due to the different behavior of the clusters during hydrolysis i.e., related to the form of the cluster in the solution [27,31]. forming a monolayer on the particle surface. Thus, almost complete saturation of the TiO 2 surface was achieved even at the lowest concentration of molybdenum cluster in solution (x = 0.1) and the following increase in the cluster concentration does not result in the increase in impregnation degree. On the contrary, partially hydrolyzed forms of tungsten cluster apparently interact with other cluster units via covalent or hydrogen bonds more effectively resulting in the growth of a "cluster shell" on the surface of the TiO 2 particles.

X-ray Photoelectron Spectroscopy (XPS)
X-ray photoelectron spectroscopy (XPS) is one of the best methods for studying the composition and the valence states of the elements in bulk materials [26,[32][33][34][35][36], and therefore it was used for investigation of n 0.1 @TiO 2 (n = 1 or 2) (Figures 2 and S4). A high noise level was observed for all cluster-related peaks, which is due to the low content of the complexes (Figure 2A,B,D,E). Nevertheless, metals peak positions, i.e., Mo3d 5/2 (228.5 eV), Mo3d 3/2 (231.7 eV), and W4d 5/2 (243.9 eV), W4d 3/2 (256.7 eV), correspond to M 2+ state, which is characteristic for cluster complexes. Iodine signals, besides main peaks at I3d 5/2 (620.1 eV) and I3d 3/2 (631.6 eV), contain shoulders at I3d 5/2 (618.3 eV) and I3d 3/2 (629.7 eV), which can be attributed to apical Iligands unsubstituted during the synthesis of initial clusters (~1 apical Iper {M 6 I 8 }) [26]. Importantly, the type of the metal in the cluster core does not affect the position of iodine ligands peaks. Spectra of titanium and oxygen are identical for both types of materials (Figures 2C,F and S4). Ti2p 3/2 and Ti2p 1/2 peaks are located at ∼458.8 and 464.5 eV, which is in agreement with the literature data for Ti 4+ in TiO 2 ( Figure 2C and Figure S4A) [32,33]. In the O1s spectrum, one can see three components located at 530.1, 531.5, and 532.7 eV ( Figures 2F and S4B). The most intensive peak (530.1 eV) can be attributed to lattice oxygen atoms in the TiO 2 [34,35]. The other two components (531.5 and 532.7 eV) can be attributed to surface groups-bridging oxygen and -OH groups, respectively [34,35]. Thus, no specific signals related to the binding of cluster complexes with surface groups of TiO 2 were observed. Nevertheless, the data obtained do not refute the formation of covalent M-O-Ti and/or hydrogen bonds M-OH/H 2 O···HO-Ti, especially taking into account the low content of the clusters in the material.  [32,33]. In the O1s spectrum, one can see three components located at 530.1, 531.5, and 532.7 eV ( Figures 2F and S4B). The most intensive peak (530.1 eV) can be attributed to lattice oxygen atoms in the TiO2 [34,35]. The other two components (531.5 and 532.7 eV) can be attributed to surface groups-bridging oxygen and -OH groups, respectively [34,35]. Thus, no specific signals related to the binding of cluster complexes with surface groups of TiO2 were observed. Nevertheless, the data obtained do not refute the formation of covalent M-O-Ti and/or hydrogen bonds M-OH/H2O···HO-Ti, especially taking into account the low content of the clusters in the material.

Surface Area and Porous Structure
Measured isotherms of nitrogen adsorption at 77 K are represented in Figure 3. All of the materials studied are possessing formally type II isotherm according to the official IUPAC classification [37], which is typical for non-porous or macroporous compounds, for which unrestricted monolayer-multilayer adsorption can occur. The inflection point (point B) lies at quite low relative pressures P/P0 (less than 2·10 −3 ) indicating the presence of micropores, though their volume is not high due to low volume adsorbed at this pressure (less than 40 mL g -1 ). At high relative pressures, hysteresis loops of H3 type are observed for all studied materials. Such a type of hysteresis is typically associated with an irregular pore network consisting of meso-and macropores that are not completely filled with condensate.

Surface Area and Porous Structure
Measured isotherms of nitrogen adsorption at 77 K are represented in Figure 3. All of the materials studied are possessing formally type II isotherm according to the official IUPAC classification [37], which is typical for non-porous or macroporous compounds, for which unrestricted monolayer-multilayer adsorption can occur. The inflection point (point B) lies at quite low relative pressures P/P 0 (less than 2·10 −3 ) indicating the presence of micropores, though their volume is not high due to low volume adsorbed at this pressure (less than 40 mL g −1 ). At high relative pressures, hysteresis loops of H3 type are observed for all studied materials. Such a type of hysteresis is typically associated with an irregular pore network consisting of meso-and macropores that are not completely filled with condensate. anomaterials 2022, 12, x FOR PEER REVIEW Figure 3. Nitrogen adsorption-desorption isotherms at 77 K. Inset shows pore sizes cording to DFT calculations. Color code: gray-TiO2, red-1 0.1 @TiO2, orange-2 0.1 @TiO The specific surface area was calculated by the conventional BET me approach. Pore volumes were calculated using the DFT method. All calculat of the porous structure are given in Table 1. The pore size distribution plot (inset in Figure 3) shows the presen mesopores with a wide maximum with a diameter of 6-8 nm. One (n 0.1 @TiO2) peaks in the microporous range in distribution graphs are attribu pores but their contribution to overall porosity is negligible (less than 0.0 about 0.4 mL g -1 ). One can note that pore volume is diminishing after impr clusters, and Vpore value of 2 0.1 @TiO2 is lower than that of 1 0.1 @TiO2, which with ICP-AES data, i.e., the higher the content of the cluster, the lower pores.

Absorption and Emission
Absorption of pure TiO2 and n 0.1 @TiO2 was studied by diffuse-reflect copy. The spectra obtained were transformed using the Kubelka-Munk fu 4A). According to the data, impregnation with cluster does not result in crease in absorption in the visible range. However, a small shoulder up to ~ forms and intensity is independent of the type of the cluster, can be obser The specific surface area was calculated by the conventional BET method and DFT approach. Pore volumes were calculated using the DFT method. All calculated parameters of the porous structure are given in Table 1. The pore size distribution plot (inset in Figure 3) shows the presence of irregular mesopores with a wide maximum with a diameter of 6-8 nm. One (TiO 2 ) or two (n 0.1 @TiO 2 ) peaks in the microporous range in distribution graphs are attributed to narrow pores but their contribution to overall porosity is negligible (less than 0.02 mL g −1 from about 0.4 mL g −1 ). One can note that pore volume is diminishing after impregnation with clusters, and V pore value of 2 0.1 @TiO 2 is lower than that of 1 0.1 @TiO 2 , which correlates well with ICP-AES data, i.e., the higher the content of the cluster, the lower the volume of pores.

Absorption and Emission
Absorption of pure TiO 2 and n 0.1 @TiO 2 was studied by diffuse-reflectance spectroscopy. The spectra obtained were transformed using the Kubelka-Munk function ( Figure 4A). According to the data, impregnation with cluster does not result in a dramatic increase in absorption in the visible range. However, a small shoulder up to~500 nm, which forms and intensity is independent of the type of the cluster, can be observed in the enlarged part of Figure 4A. Despite the different content of the complexes, the overall impact of cluster impregnation on the absorption of the materials is the same. This is due to the difference in absorption of the clusters themselves-molybdenum clusters have wider absorption than tungsten ones, but a smaller degree of {Mo 6 I 8 } impregnation equalizes the overall effect. In turn, the effect of cluster content is appearing in the UV region, since 2 0.1 @TiO 2 has higher than 1 0.1 @TiO 2 absorption in 250-350 nm range.  To measure optical band gap (Eg), the Tauc plot was used. The dependence of (αhν) versus hν was plotted and interception of straight lines with y = 0 gives Eg values ( Figure 4B). Despite the admixture of brookite phase, pure TiO2 and both cluster-containing materials demonstrates typical for anatase band gap value of 3.2 eV [17]. Interestingly, but on the spectra of both n 0.1 @TiO2 one can see another straight line with Eg of 1.8 eV, which can be attributed to the shoulder in the absorption spectra ( Figure 4A). This value can be referred to hydrolyzed cluster complex on the surface of the particles (e.g., Eg of [{Mo6I8}(H2O)2(OH)4]·2H2O is 1.7 eV [38]), thus confirming our earlier suggestion.
Emission spectroscopy is useful tool to estimate the separation efficiency and transfer ability of photogenerated charge carriers since electron-hole pairs recombination results in emission. Emission spectra of pure TiO2 and n 0.1 @TiO2 were recorded for the samples in solid state ( Figure 4C). Both n 0.1 @TiO2 demonstrates similar intensity of the emission which is lower than the intensity of pure TiO2 for ~25%. Lowering of the emission intensity indicates higher lifetime of electron-hole pairs. Note, that there is no cluster originated emission on the spectra indicating effective charge carriers transfer between complex and matrix.

Photocatalytic Degradation of RhB
UV-light. The photocatalytic activity of the samples was evaluated under UV-light irradiation (λ = 365 nm, 13 mW cm -2 ). Rhodamine B (RhB) was chosen as the model organic pollutant ( Figure S5). Before carrying out photocatalytic experiments, the reaction mixture was kept stirred in dark for 2 h to achieve sorption-desorption equilibrium. After that, the mixture was irradiated with UV light under constant stirring, and aliquots were taken in different time intervals. The dye degradation rate was monitored on UV-vis spectra by reduction in characteristic RhB band at 554 nm ( Figure 5). Effective rate constants were calculated from linear approximation of ln(C/C0) vs. t plots (where C is concentration of RhB at corresponding t, C0 is initial concentration of RhB, t is time when aliquots were taken) (inserts in Figure 5) and presented in Table 2. According to the data obtained, pure TiO2 does not degrade dye completely even after 90 min of irradiation and its keff is 0.02 min -1 . In turn, both n 0.1 @TiO2 demonstrate close activity completely decomposing RhB after 45 min with keff of ~0.099-0.11 min -1 , which is ~5 times higher than that of titania nanoparticles. Taking into account experimental conditions, rate constant values of n 0.1 @TiO2 are comparable to the most efficient photocatalysts known in literature [39,40]. Concerning the dependency of n x @TiO2 efficiency on cluster content, one can see, that keff behaves differently for materials containing molybdenum or tungsten clusters ( Table 2). All of the 1 x @TiO2 demonstrate similar activity with keff of ~0.1 min -1 , while for 2 x @TiO2 keff gradually decreases from 0.11 (for x = 0.1) to 0.025 (for x = 1.5) min -1 . This tendency correlates well To measure optical band gap (E g ), the Tauc plot was used. The dependence of (αhν) 1 2 versus hν was plotted and interception of straight lines with y = 0 gives E g values ( Figure 4B). Despite the admixture of brookite phase, pure TiO 2 and both cluster-containing materials demonstrates typical for anatase band gap value of 3.2 eV [17]. Interestingly, but on the spectra of both n 0.1 @TiO 2 one can see another straight line with E g of 1.8 eV, which can be attributed to the shoulder in the absorption spectra ( Figure 4A). This value can be referred to hydrolyzed cluster complex on the surface of the particles (e.g., E g of [{Mo 6

I 8 }(H 2 O) 2 (OH) 4 ]·2H 2 O is 1.7 eV [38]), thus confirming our earlier suggestion.
Emission spectroscopy is useful tool to estimate the separation efficiency and transfer ability of photogenerated charge carriers since electron-hole pairs recombination results in emission. Emission spectra of pure TiO 2 and n 0.1 @TiO 2 were recorded for the samples in solid state ( Figure 4C). Both n 0.1 @TiO 2 demonstrates similar intensity of the emission which is lower than the intensity of pure TiO 2 for~25%. Lowering of the emission intensity indicates higher lifetime of electron-hole pairs. Note, that there is no cluster originated emission on the spectra indicating effective charge carriers transfer between complex and matrix.

Photocatalytic Degradation of RhB
UV-light. The photocatalytic activity of the samples was evaluated under UV-light irradiation (λ = 365 nm, 13 mW cm −2 ). Rhodamine B (RhB) was chosen as the model organic pollutant ( Figure S5). Before carrying out photocatalytic experiments, the reaction mixture was kept stirred in dark for 2 h to achieve sorption-desorption equilibrium. After that, the mixture was irradiated with UV light under constant stirring, and aliquots were taken in different time intervals. The dye degradation rate was monitored on UV-vis spectra by reduction in characteristic RhB band at 554 nm ( Figure 5). Effective rate constants were calculated from linear approximation of ln(C/C 0 ) vs. t plots (where C is concentration of RhB at corresponding t, C 0 is initial concentration of RhB, t is time when aliquots were taken) (inserts in Figure 5) and presented in Table 2. According to the data obtained, pure TiO 2 does not degrade dye completely even after 90 min of irradiation and its k eff is 0.02 min −1 . In turn, both n 0.1 @TiO 2 demonstrate close activity completely decomposing RhB after 45 min with k eff of~0.099-0.11 min −1 , which is~5 times higher than that of titania nanoparticles. Taking into account experimental conditions, rate constant values of n 0.1 @TiO 2 are comparable to the most efficient photocatalysts known in literature [39,40]. Concerning the dependency of n x @TiO 2 efficiency on cluster content, one can see, that k eff behaves differently for materials containing molybdenum or tungsten clusters ( Table 2). All of the 1 x @TiO 2 demonstrate similar activity with k eff of~0.1 min −1 , while for 2 x @TiO 2 k eff gradually decreases from 0.11 (for x = 0.1) to 0.025 (for x = 1.5) min −1 . This tendency correlates well with the real cluster content determined using ICP-AES ( Figure 1H), i.e., in 1 x @TiO 2 the content of {Mo 6 I 8 } is independent on x, thus here we have similar activity of all of the samples. In turn, the content of {W 6 I 8 } in 2 x @TiO 2 increases with the increase in the amount of complex in reaction mixture, and k eff depends linearly on cluster core content ( Figure S6). This behavior is probably associated with shielding of the titanium dioxide surface when a large amount of the complex is deposited, due to which TiO 2 does not participate in the photocatalytic reaction, thus the material exhibits mostly the activity of cluster. This phenomenon demonstrates well the synergistic effect in the titled system. The effect of catalyst concentration was studied using 10 or 30 mg of n 0.1 @TiO 2 per 80 mL of the reaction mixture (C cat are 0.125 or 0.375 g L −1 ). According to the data obtained, the changes in catalyst concentrations resulted in a decrease in efficiency (Table 2), which makes a concentration of 0.25 g L −1 the most optimal. Since the highest catalytic efficiency under UV light along with the lowest cluster content was shown for n 0.1 @TiO 2 , these samples were chosen for further studies. samples. In turn, the content of {W6I8} in 2 x @TiO2 increases with the increase in the amount of complex in reaction mixture, and keff depends linearly on cluster core content ( Figure S6). This behavior is probably associated with shielding of the titanium dioxide surface when a large amount of the complex is deposited, due to which TiO2 does not participate in the photocatalytic reaction, thus the material exhibits mostly the activity of cluster. This phenomenon demonstrates well the synergistic effect in the titled system. The effect of catalyst concentration was studied using 10 or 30 mg of n 0.1 @TiO2 per 80 mL of the reaction mixture (Ccat are 0.125 or 0.375 g L -1 ). According to the data obtained, the changes in catalyst concentrations resulted in a decrease in efficiency (Table 2), which makes a concentration of 0.25 g L -1 the most optimal. Since the highest catalytic efficiency under UV light along with the lowest cluster content was shown for n 0.1 @TiO2, these samples were chosen for further studies.

Sample
Rate Constant (keff), min -1 TiO2 0.02 n = 1 n = 2 n 0.1 @TiO2 0.099 0.11 n 0.5 @TiO2 0.10 0.075 n 1 @TiO2 0.11 0.062 n 1.5 @TiO2 0.096 0.025 Ccat = 0.125 g L -1 n = 1 n = 2 n 0.1 @TiO2 0.055 0.026 Ccat = 0.375 g L -1 n = 1 n = 2 n 0.1 @TiO2 0.047 0.081 Cyclic experiments. To study the stability of the materials in photocatalytic reactions cyclic experiments were conducted. n 0.1 @TiO2 samples were chosen due to highest activity at lowest cluster content. In general, 5 runs of irradiation were performed. To avoid the accumulation of oxidized byproducts, which can decrease the rate of the reaction, irradiation time was increased to 60 min. After each run, the aliquot was taken and the concentration of RhB and n 0.1 @TiO2 was adjusted to the initial state. The results obtained are presented in Figure 6, and one can see, that both of the samples do not loss their activity up to 5 cycles.   Cyclic experiments. To study the stability of the materials in photocatalytic reactions cyclic experiments were conducted. n 0.1 @TiO 2 samples were chosen due to highest activity at lowest cluster content. In general, 5 runs of irradiation were performed. To avoid the accumulation of oxidized byproducts, which can decrease the rate of the reaction, irradiation time was increased to 60 min. After each run, the aliquot was taken and the concentration of RhB and n 0.1 @TiO 2 was adjusted to the initial state. The results obtained are presented in Figure 6, and one can see, that both of the samples do not loss their activity up to 5 cycles.  (Figure 7). Similar to experiments under UV-light, both n 0.1 @TiO2 demonst close activity decomposing RhB after ~45-50 min with keff of 0.12 and 0.14 min -1 for and 2 correspondingly. One can see, that all of the samples show some increase i activity, which is due to the continuous spectrum of solar light providing more pho with a suitable wavelength.

Scavengers
Scavengers are compounds that can selectively react with certain active species OH • , O2 •− , e − , or h + . Decrease in the rate of the photocatalytic reaction in the presence of compounds indicates the impact of specific active species, which is helpful for revealing m anism of photocatalytic activity. Here we used i PrOH, AgNO3, and Na2C2O4 as specific engers for OH • , e − , and h + , respectively. To study the effect of O2 •the reaction mixture deaerated by the bubbling of Ar gas. The photocatalytic experiments were conducted a  (Figure 7). Similar to experiments under UV-light, both n 0.1 @TiO 2 demonstrated close activity decomposing RhB after~45-50 min with k eff of 0.12 and 0.14 min −1 for n = 1 and 2 correspondingly. One can see, that all of the samples show some increase in the activity, which is due to the continuous spectrum of solar light providing more photons with a suitable wavelength.  (Figure 7). Similar to experiments under UV-light, both n 0.1 @TiO2 demonstrated close activity decomposing RhB after ~45-50 min with keff of 0.12 and 0.14 min -1 for n = 1 and 2 correspondingly. One can see, that all of the samples show some increase in the activity, which is due to the continuous spectrum of solar light providing more photons with a suitable wavelength.

Scavengers
Scavengers are compounds that can selectively react with certain active species, e.g., OH • , O2 •− , e − , or h + . Decrease in the rate of the photocatalytic reaction in the presence of such compounds indicates the impact of specific active species, which is helpful for revealing mechanism of photocatalytic activity. Here we used i PrOH, AgNO3, and Na2C2O4 as specific scavengers for OH • , e − , and h + , respectively. To study the effect of O2 •the reaction mixture was deaerated by the bubbling of Ar gas. The photocatalytic experiments were conducted as earlier and keff in the presence of scavengers were calculated ( Figure 8A,B, Table S2). Using obtained keff values, corresponding relative activities for each type of the scavenger were calculated according to Equation (2) ( Figure 8C).

Scavengers
Scavengers are compounds that can selectively react with certain active species, e.g., OH • , O 2 •− , e − , or h + . Decrease in the rate of the photocatalytic reaction in the presence of such compounds indicates the impact of specific active species, which is helpful for revealing mechanism of photocatalytic activity. Here we used i PrOH, AgNO 3 , and Na 2 C 2 O 4 as specific scavengers for OH • , e − , and h + , respectively. To study the effect of O 2 •− the reaction mixture was deaerated by the bubbling of Ar gas. The photocatalytic experiments were conducted as earlier and k eff in the presence of scavengers were calculated ( Figure 8A,B, Table S2). Using obtained k eff values, corresponding relative activities for each type of the scavenger were calculated according to Equation (2) ( Figure 8C). One can see that all of the scavengers used in certain degree reduce the activity of the catalysts, indicating impact of all active species on photocatalytic process. The lowest effect was achieved in the presence of i PrOH (decrease in activity for 11 and 13% for 1 0.1 @TiO2 and 2 0.1 @TiO2 correspondingly), therefore OH • is the least active particle for both type of catalysts. Scavenging of e − and h + resulted in greater activity lowering-for 81 and 72% for 1 0.1 @TiO2 and for 67 and 57% for 2 0.1 @TiO2. The highest inhibition was achieved in deaerated atmosphere (decrease in activity for 85 and 91% for 1 0.1 @TiO2 and 2 0.1 @TiO2 correspondingly) proving that⋅O2 •are the dominant active species in photodegradation of RhB.

Photocatalytic Mechanism
Thus, participation of all active species -h + , e − , OH • and O2 •− , in photocatalytic process along with significantly higher activity of the material compared with pure TiO2 allow us to suggest the formation of S-scheme (also known as direct Z-scheme) heterojunction in n 0.1 @TiO2 [41]. In such systems, due to the formation of heterojunction, resulting in band curving, and activity of both components under irradiation, less active electrons and holes can recombine thereby preserving electrons and holes having higher energy. In order to illustrate band positions in the materials, EVB-EF (EVB is the edge positions of valence bands, EF is Fermi level potential) distances in n 0.1 @TiO2 were determined using valence band XPS (VB-XPS) ( Figure 9). One can see on the Figure 9A, that intercept of straight lines gives us value of 2.86 eV, which is in good agreement with literature data for TiO2 [42]. Nevertheless, similar to absorption graphs, there are additional shoulders, which gives EVB-EF distances of 0.25 eV independently to the type of the metal in cluster core ( Figure  9B). These values can be attributed to the presence of clusters in the materials.  One can see that all of the scavengers used in certain degree reduce the activity of the catalysts, indicating impact of all active species on photocatalytic process. The lowest effect was achieved in the presence of i PrOH (decrease in activity for 11 and 13% for 1 0.1 @TiO 2 and 2 0.1 @TiO 2 correspondingly), therefore OH • is the least active particle for both type of catalysts. Scavenging of e − and h + resulted in greater activity lowering-for 81 and 72% for 1 0.1 @TiO 2 and for 67 and 57% for 2 0.1 @TiO 2 . The highest inhibition was achieved in deaerated atmosphere (decrease in activity for 85 and 91% for 1 0.1 @TiO 2 and 2 0.1 @TiO 2 correspondingly) proving that·O 2 •− are the dominant active species in photodegradation of RhB.

Photocatalytic Mechanism
Thus, participation of all active species-h + , e − , OH • and O 2 •− , in photocatalytic process along with significantly higher activity of the material compared with pure TiO 2 allow us to suggest the formation of S-scheme (also known as direct Z-scheme) heterojunction in n 0.1 @TiO 2 [41]. In such systems, due to the formation of heterojunction, resulting in band curving, and activity of both components under irradiation, less active electrons and holes can recombine thereby preserving electrons and holes having higher energy. In order to illustrate band positions in the materials, E VB -E F (E VB is the edge positions of valence bands, E F is Fermi level potential) distances in n 0.1 @TiO 2 were determined using valence band XPS (VB-XPS) ( Figure 9). One can see on the Figure 9A, that intercept of straight lines gives us value of 2.86 eV, which is in good agreement with literature data for TiO 2 [42]. Nevertheless, similar to absorption graphs, there are additional shoulders, which gives E VB -E F distances of 0.25 eV independently to the type of the metal in cluster core ( Figure 9B). These values can be attributed to the presence of clusters in the materials. One can see that all of the scavengers used in certain degree reduce the activity of the catalysts, indicating impact of all active species on photocatalytic process. The lowest effect was achieved in the presence of i PrOH (decrease in activity for 11 and 13% for 1 0.1 @TiO2 and 2 0.1 @TiO2 correspondingly), therefore OH • is the least active particle for both type of catalysts. Scavenging of e − and h + resulted in greater activity lowering-for 81 and 72% for 1 0.1 @TiO2 and for 67 and 57% for 2 0.1 @TiO2. The highest inhibition was achieved in deaerated atmosphere (decrease in activity for 85 and 91% for 1 0.1 @TiO2 and 2 0.1 @TiO2 correspondingly) proving that⋅O2 •are the dominant active species in photodegradation of RhB.

Photocatalytic Mechanism
Thus, participation of all active species -h + , e − , OH • and O2 •− , in photocatalytic process along with significantly higher activity of the material compared with pure TiO2 allow us to suggest the formation of S-scheme (also known as direct Z-scheme) heterojunction in n 0.1 @TiO2 [41]. In such systems, due to the formation of heterojunction, resulting in band curving, and activity of both components under irradiation, less active electrons and holes can recombine thereby preserving electrons and holes having higher energy. In order to illustrate band positions in the materials, EVB-EF (EVB is the edge positions of valence bands, EF is Fermi level potential) distances in n 0.1 @TiO2 were determined using valence band XPS (VB-XPS) ( Figure 9). One can see on the Figure 9A, that intercept of straight lines gives us value of 2.86 eV, which is in good agreement with literature data for TiO2 [42]. Nevertheless, similar to absorption graphs, there are additional shoulders, which gives EVB-EF distances of 0.25 eV independently to the type of the metal in cluster core ( Figure  9B). These values can be attributed to the presence of clusters in the materials.  According to literature data, the valence band potential (E VB ) of TiO 2 is −7.55 eV vs. E VAC (3.05 eV vs. NHE) [43][44][45]. Thus, the conduction band potential (E CB ) of titania was calculated using equation E g = E VB − E CB and was found to be −4.35 eV vs. E VAC (0.05 eV vs. NHE). Fermi level position (E F ) was calculated using E VB -E F distance (2.86 eV) and was found to be −4.69 eV vs. E VAC (0.19 eV vs. NHE). Assuming that when the semiconductors are in contact, their Fermi levels are aligned, while the positions of the valence and conduction bands (VB and CB) remain unchanged, we calculated E VB of clusters, which are equal to −4.94 eV vs. E VAC (0.54 eV vs. NHE), and then the position of the conduction band-−3.14 eV vs. E VAC (−1.36 eV vs. NHE). Therefore, using the data obtained we can propose the band gap structure and mechanism of the RhB photodegradation ( Figure 10). Thus, according to calculations, the potentials obtained are sufficient for the generation of all radicals, which is in agreement with scavenger experiments and overall confirms the formation of S-scheme heterojunctions.
Nanomaterials 2022, 12, x FOR PEER REVIEW According to literature data, the valence band potential (EVB) of TiO2 is -7.5 EVAC (3.05 eV vs. NHE) [43][44][45]. Thus, the conduction band potential (ECB) of tita calculated using equation E = E − E and was found to be -4.35 eV vs. EVAC vs. NHE). Fermi level position (EF) was calculated using EVB-EF distance (2.86 eV) found to be -4.69 eV vs. EVAC (0.19 eV vs. NHE). Assuming that when the semicon are in contact, their Fermi levels are aligned, while the positions of the valence a duction bands (VB and CB) remain unchanged, we calculated EVB of clusters, w equal to −4.94 eV vs. EVAC (0.54 eV vs. NHE), and then the position of the con band--3.14 eV vs. EVAC (-1.36 eV vs. NHE). Therefore, using the data obtained propose the band gap structure and mechanism of the RhB photodegradation ( Fig  Thus, according to calculations, the potentials obtained are sufficient for the gene all radicals, which is in agreement with scavenger experiments and overall conf formation of S-scheme heterojunctions. The mechanism of dye photodegradation can be depicted according to Equat (7). First, both components absorb photons producing electron-hole pairs (Equat Next, the electron on TiO2 and the hole on the cluster, having low potentials, rec (Equation (4)). In turn, the electron on the cluster and the hole on TiO2, preserve recombination of other pair, have sufficient potential to produce O2 •or OH • (Eq (5) and (6)). Then all active species formed mineralize RhB molecule (Equation

Conclusions
To conclude, anatase-brookite mixed-phase TiO 2 nanoparticles were obtained by an aqueous hydrolysis of titanium (IV) isopropoxide in a large excess of water. Impregnation of the titania with a varied amount of [{M 6 I 8 }(DMSO) 6 ](NO 3 ) 4 (M = Mo (1), W (2)) resulted in the formation of hybrid cluster-containing photocatalysts n x @TiO 2 (n is Mo or W cluster, x is the amount of cluster used in impregnation reaction). Analysis of the composition revealed a strong effect of the type of the cluster on impregnation rate: more hydrolytically unstable complex 1 impregnates TiO 2 surface in equal amount regardless of x, while in the case of more stable complex 2 the content of {W 6 I 8 } grows linearly with x. Cluster deposition results in moderate increase in the absorption in visible range up to~500 nm and appearance of additional cluster-related band gap with value of 1.8 eV. Photocatalytic activity of the samples in RhB degradation is directly affected by the amount of the cluster deposited on the particle-all 1 x @TiO 2 demonstrates similar activity, and for 2 x @TiO 2 decrease in the efficiency along with x was observed. Thus, highest efficiency was achieved for n 0.1 @TiO 2 with rate constants of~0.1-0.14 min −1 under both UV-and sunlight irradiation, which is 4-5 times higher than those of pure TiO 2 . Both the materials are stable for at least for 5 cycles of photocatalytic degradation of the dye without any loss in the efficiency. High activity of the materials comparing to pure TiO 2 along with impact of all of the active species, revealed in scavengers' experiments, indicates formation of S-scheme heterojunction catalysts. Thus, the materials obtained here demonstrate high efficiency both under UV and sunlight irradiation and, therefore, they can be used for efficient wastewater treatment from organic pollutants.

Supplementary Materials:
The following supporting information can be downloaded at: https:// www.mdpi.com/article/10.3390/nano12234282/s1, Figure S1: Representative structure of [{M 6 I 8 }L 6 ] n (M = Mo or W; green octahedron is M 6 , violet spheres are inner iodine ligands, blue speres are apical ligands (L) of any nature, n-charge) units; Figure S2: XRD patterns of n x @TiO 2 , n = 1 (A) or 2 (B) in comparison with pure TiO 2 and calculated anatase and brookite diffractograms; Figure S3: FTIR spectra of pure TiO 2 and n x @TiO 2 , n = 1 (A) or 2 (B); Figure S4: High-resolution XPS spectra of Ti2p (A) and O1s (B) core levels in 2 0.1 @TiO 2 ; Figure S5: Structure of RhB molecule; Figure S6: The dependency of k eff of 2 x @TiO 2 on real content of {W 6 I 8 } determined using ICP-AES; Table S1: Determination of the amount of {M 6 I 8 } units anchored on TiO 2 using ICP-AES; Table S2: Effective rate constants (k eff ) and R 2 values of RhB decomposition by n x @TiO 2 in the presence of different scavengers.

Data Availability Statement:
The data that support the findings of this study are available from the corresponding author upon reasonable request.