Elucidating the Photocatalytic Behavior of TiO2-SnS2 Composites Based on Their Energy Band Structure

TiO2-SnS2 composite semiconducting photocatalysts with different building component ratios were prepared by hydrothermal synthesis (TiO2-SnS2-HT) and by immobilization of commercial TiO2 and SnS2 particles (TiO2-SnS2-COMM). The band gap values, which determine the catalysts’ photoactivity, were examined by diffuse reflectance spectroscopy and Kubelka–Munk transformations. The catalysts’ surface properties: specific surface area, charge and adsorption capacitance at the solid–solution interface were characterized using BET analysis, potentiometric titration and electrochemical impedance spectroscopy, respectively. The electronic band structure of TiO2-SnS2 photocatalyst, as the key property for the solar-driven photocatalysis, was deduced from the thermodynamic data and the semiconducting parameters (type of semiconductivity, concentration of the charge carriers, flat band potential) obtained by Mott–Schottky analysis. The photoactivity of both composites was studied in photocatalytic treatment of diclofenac (DCF) under simulated solar irradiation and was compared to the benchmark photocatalyst (TiO2 P25) activity. The influence of process parameters, such as pH, H2O2, and composite formulation on the effectiveness of DCF removal and conversion was investigated and discussed by employing response surface modeling (RSM) approach. The photocatalytic efficiency of both composite materials was discussed on the basis of the hetereojunction formation that facilitated the photoelectron transfer, promoting more efficient photocatalytic degradation of DCF.


Photocatalyst Synthesis and Immobilization
Hydrothermal method, according to procedure adopted from Zhang et al. [9], was applied to prepare TiO 2 , SnS 2 and their composites denoted as TiO 2 -SnS 2 -HT. TiO 2 was synthesized by dissolving an aliquot of TBO precursor in a 5% v/v solution of acetic acid in ethanol with a constant stirring in Teflon reaction vessel, which was then transferred to a stainless steel autoclave, where treated for 12 h at 180 • C. After cooling naturally to room temperature, the obtained suspension was rinsed with distilled water, centrifuged (3500 rpm for 3 min), dried in a vacuum (3 h at 60 • C), and then homogenized with a porcelain pestle and mortar. The same procedure was applied for SnS 2 and TiO 2 -SnS 2 -HT composite synthesis, using corresponding precursors (tin(IV) chloride, thioacetamide, and TBO in later case), while their stoichiometric were varied to obtain the composite with different SnS 2 content (wt % of 5, 27.5, 50, 72.5). Composites denoted as TiO 2 -SnS 2 -COMM were prepared from the commercial TiO 2 and SnS 2 particles during immobilization procedure, whereas their weights were varied to obtain abovementioned ratios.
The immobilization procedure included the addition of as-prepared composite or a mixture of commercial particles into titania/silica binder sol (TSB) to obtain TiO 2 -SnS 2 -HT and TiO 2 -SnS 2 -COMM, respectively. TSB was made by mixing two as-prepared sols (nanocrystaline titania sol synthetized from TTIP and silica sol synthetized from TEOS), colloidal SiO 2 and ethanol [28]. The suspended particles within TSB sol were firstly homogenized using ultrasonic bath, and then the final sol suspension was immobilized on (i) round titanium discs (d = 1 mm, r = 12 mm) in three layers, and (ii) round glass plates (r = 35.5 mm) in one layer using spin coating (1500 rpm) technique with KW-4A Spin Coater (Chemat Technology, Northridge, CA, USA). The final fixation step was performed by heating the coated titanium discs or glass plates in a laboratory oven (2 h at 200 • C).

Photocatalyst Characterization
Diffuse reflectance spectra (DRS) of immobilized photocatalysts, were measured using UV-vis spectrophotometer equipped with an integrating sphere, Lambda 650S (Perkin Elmer, Waltham, MA, USA). The acquired reflectance vs. wavelength spectra were transformed into the Kubelka-Munk function (KM) vs. photon energy (hν) in order to obtain the band gap values [29][30][31]. Nitrogen adsorption analysis, performed using Gemini 2380 instrument (Micromeritics, Norcross, GA, USA), was used to determine the specific surface area of composites and their pure components. Brunauer-Emmett-Teller (BET) surface area was calculated from BET plot. The samples of immobilized composites were obtained by carefully scratching the films and collecting the corresponding powder, which was then grounded in an agate mortar. The same procedure for obtaining such powdered samples of composites was applied prior determination of point of zero charge (pH PCZ ) values. Handylab pH/LF portable pH meter (Schott Instruments GmbH, Mainz, Germany) was used for pH monitoring during potentiometric titration applied for the determination of pH PCZ of TiO 2 -SnS 2 composites. The pH PCZ values were determined according to the modified procedure of Uppal et al. [32], while detail description of applied methodology is provided in the previous study [33].
Electrochemical characterization of photocatalysts immobilized on titanium substrates was performed using Solartron potentiostat/galvanostat 1287 with FRA 1260 in a conventional three-electrode cell: Ti-coated disc was working electrode, the counter electrode was Pt and the reference electrode was Ag|AgCl (E = 0.208 V vs. SHE). The electrolyte was 3% NaCl solution, pure or spiked with 0.1 mM DCF. The structure of catalyst films-electrolyte interface was investigated at the open circuit potential (E ocp ) using electrochemical impedance spectroscopy (EIS); frequency ranged from 100 kHz to 5 mHz at an ac voltage amplitude of ±5 mV. The experimental data were fitted using the complex non-linear least squares fit analysis software [34] and values of elements of proposed electric equivalent circuit (EEC) were derived with χ 2 values <5 × 10 −3 . The electronic-semiconducting properties of catalyst films were investigated by Mott-Schottky analysis [22][23][24]. The interfacial capacitance values were obtained from EIS measurements as a function of both frequency and applied potential in a rapid polarization scan. Detailed description is given in Supplementary Material.

Photocatalytic Activity under Solar Irradiation
The photocatalytic experiments were performed in a water-jacket batch reactor (V = 0.09 L and T = 25.0 ± 0.2 • C). The source of simulated solar irradiation was a 450 W Xenon arc lamp (Osram, Munich, Germany) situated in Oriel/Newport, USA housing with collimating optics. An Oriel AM1.5 G air mass filter was situated in the path of the collimated beam, mimicking solar spectral characteristics when the Sun is at a zenith angle of 48.2 • . The light intensity, 124.78 ± 0.11 mW cm −2 , was determined by pyranometer CMP21 (Kipp & Zonen, Delft, The Netherlands), while the intensity of UV-A irradiation emitted by solar simulator was found to be 2.05 ± 0.07 mW cm −2 , determined using UVX radiometer equipped with UVX-36 longwave sensor (both UVP, Cambridge, UK). Aqueous solution of DCF (c 0 = 0.1 mM) was treated by solar driven photocatalytic processes using TiO 2 -SnS 2 -COMM or TiO 2 -SnS 2 -HT in the presence and absence of H 2 O 2, whereas TiO 2 P25 was used as a benchmark material. The DCF solution with adjusted pH was then spiked with H 2 O 2 (where applicable) according to full factorial (FFD) or Box-Behnken designs (BBD) ( Table 1 and Tables S1-S4, Supplementary Material), which was followed by the immersion of the glass plates into the reactor placed on an orbital shaker DOS-20 (90 rpm, neoLab, Heidelberg, Germany). Adsorption equilibrium was reached within 30 min in a dark, and thereafter the reaction solution was exposed to simulated solar irradiation. The samples (500 µL aliquots) were taken during experiments at −30 (30 min prior to irradiation), 0 (the start of irradiation), 15, 30, 45, and 60 min, filtered using Chromafil XTRA RC (25 mm, 0.45 µm, Macherey Nagel, Düren, Germany), quenched with CH 3 OH and submitted to HPLC analysis. All experiments were repeated at least three times and averages are reported; the reproducibility was >97.3%. DCF concentration was monitored by HPLC analysis, using Series 10 apparatus equipped with UV-DAD, SPD-M10A VP (Shimadzu, Japan), Nucleosil C18 column (5 µm, 25.0 cm × 4.6 mm, Macherey Nagel, Düren, Germany), and mobile phase CH 3 OH 0.1% formic acid operating at 1.0 mL min −1 flow. UV-vis spectrophotometer, Lambda EZ 201, Perkin Elmer (USA) was used for the monitoring of H 2 O 2 by metavanadate method [35]. In desorption tests, the plates used in photocatalytic experiments were immersed in NaOH solution (90 mL, pH 8.00 ± 0.05) and placed on the shaker for 30 min. The solution was analyzed according to the above procedure. The influence of initial pH, H 2 O 2 , and SnS 2 wt % within TiO 2 -SnS 2 composites, on DCF removal and conversion by solar driven photocatalytic treatment was correlated by means of RSM. Related calculations, procedure, and analysis are described in Supplementary Material.

Semiconducting and Surface Properties of TiO 2 -SnS 2 Composite Materials
DRS analysis of TiO 2 -SnS 2 -COMM and TiO 2 -SnS 2 -HT composites, with different SnS 2 content ranging from 5 to 72.5 wt %, was performed to determine the band gap values, E g of composite materials studied. DRS spectra of the pure components TiO 2 -HT, SnS 2 -HT and SnS 2 MKN-900 were also recorded. The spectra obtained ( Figure S1A,B, Supplementary Material) were transformed using Kubelka-Munk function, Figure 1A,B. The E g values determined are given in Table 2. While F(R) 1/2 vs. hν plots of TiO 2 -SnS 2 -HT are characterized with only one flat section, the existence of two flat sections in plots of TiO 2 -SnS 2 -COMM, more emphasized with SnS 2 wt % increase, point to the segregation of TiO 2 and SnS 2 composite phases. Such results are in accordance with the established morphology revealed by SEM/EDX analysis showing phase separation in the COMM composite while the HT composite consisted of TiO 2 matrix with embedded and homogenously dispersed SnS 2 small proto-platelets [33,36]. The described morphology characteristics, i.e., phase coupling degree of composite components, enable/disable charge carries transfer through both materials present within. The band gap values of both composites decreased with SnS 2 wt % increase. However, this effect became less significant upon exceeding SnS 2 wt % over 50%, Table 2. Based on the results obtained, the composites possessing SnS 2 wt % in range 5-50% were set as preferable to be studied in the further tests. Accordingly, the formulation of 72.5 wt % of TiO 2 and 27.5 wt % of SnS 2 presents the middle point of such set range.   * Adopted from previous study [5].
Semiconducting properties of TiO 2 -SnS 2 -COMM and TiO 2 -SnS 2 -HT films on titanium substrate were further probed by Mott-Schottky analysis. This analysis was performed taking the middle point formulation (27.5 wt % of SnS 2 ) for both composites. Mott-Schottky plots for composites ( Figure 2) show linear dependence of space charge capacitance (C −2 SC ) against applied potential (E) with a positive slope suggesting n-type semiconducting behavior for both composite films. Under depletion conditions, Mott-Schottky approximation of the capacitance-voltage relationship for n-type semiconductor has the form [37] where ε 0 is vacuum permittivity, ε is dielectric constant of composite film, e is electron charge, N D is donor concentration, E FB is flat band potential, k is Boltzmann constant, and T is thermodynamic temperature.
where pH pzc is the point of zero charge, E 0 F is Fermi level of semiconductor at pH pzc , and ∆φ H is the potential drop in Helmholtz double layer at the solid|electrolyte solution interface. At a certain potential, the Fermi level lies at the same energy as the electrolyte solution potential. There is no net transfer of charge, and hence there is no band bending. This potential is therefore referred to as the flat band potential, E FB .
The specific surface area of studied TiO 2 -SnS 2 composites and their pure components (TiO 2 P25, SnS 2 MKN-900, TiO 2 -HT, and SnS 2 -HT) was determined using BET analysis. The surface charge of studied composites was evaluated by means of their pH PZC values, determined by potentiometric titration. The results are summarized in Table 3 (for composites) and Table S5 (Supplementary Material) (for their pure Components), and presented graphically for pH PZC of both composites ( Figure S2, Supplementary material). BET surface areas of studied composites differ significantly (Table 3). Specific area of TiO 2 -SnS 2 -HT is almost twice as large as that of TiO 2 -SnS 2 -COMM. A similar trend in the specific surface area values observed for their pure components was assigned to the particle size and crystalline phase types (Table S5). TiO 2 -HT is characterized by smaller particles comparing to TiO 2 P25 and exclusively anatase crystalline phase [33,36]. Both SnS 2 materials possess the same crystalline phase (berndtite), exhibiting rather different morphologies with SnS 2 -HT comprised of significantly smaller structures in comparison to SnS 2 MKN-900 [33]. Taking into account the pH PZC values determined, it can be concluded that as-synthetized particles have significantly lower pH PZC values than commercial ones (Table 3), which might be reflected in their photocatalytic activity, as investigated further in the study.
semiconductor [38,39] = − + where pHpzc is the point of zero charge, 0 F E is Fermi level of semiconductor at pHpzc, and H   is the potential drop in Helmholtz double layer at the solid|electrolyte solution interface. At a certain potential, the Fermi level lies at the same energy as the electrolyte solution potential. There is no net transfer of charge, and hence there is no band bending. This potential is therefore referred to as the flat band potential, EFB. The specific surface area of studied TiO2-SnS2 composites and their pure components (TiO2 P25, SnS2 MKN-900, TiO2-HT, and SnS2-HT) was determined using BET analysis. The surface charge of studied composites was evaluated by means of their pHPZC values, determined by potentiometric titration. The results are summarized in Table 3 (for composites) and Table S5 (Supplementary Material) (for their pure Components), and presented graphically for pHPZC of both composites ( Figure S2, Supplementary material). BET surface areas of studied composites differ significantly (Table 3). Specific area of TiO2-SnS2-HT is almost twice as large as that of TiO2-SnS2-COMM. A similar trend in the specific surface area values observed for their pure components was assigned to the particle size and crystalline phase types (Table S5). TiO2-HT is characterized by smaller particles comparing to TiO2 P25 and exclusively anatase crystalline phase [33,36]. Both SnS2 materials possess the same crystalline phase (berndtite), exhibiting rather different morphologies with SnS2-HT comprised of significantly smaller structures in comparison to SnS2 MKN-900 [33]. Taking into account the pHPZC values determined, it can be concluded that as-synthetized particles have significantly lower pHPZC values than commercial ones (Table 3), which might be reflected in their photocatalytic activity, as investigated further in the study. Table 3. Surface properties (specific surface area and point of zero charge values) of studied TiO2-SnS2 composites (72.5:27.5 wt %).

Material
BET Surface Area, m 2 g −1 pHPZC TiO2-SnS2-COMM 96.00 ± 0.78 6.31 TiO2-SnS2-HT 160.58 ± 0.46 4.61  The further characterization of composites' surface properties was conducted by EIS measurements, performed at E ocp representing real conditions of potential catalyst application. The impedance spectra of both composite films on Ti plates with and without presence of the organic pollutant DCF (Figure 3), were fitted using EEC (insert in Figure 3); corresponding EEC parameter values are summarized in Table 4. R Ω corresponds to the ohmic (electrolyte) resistance. The R ct and C dl parameters can be linked to the charge transfer processes that occur in an outer porous part of the composite catalyst film at the film-electrolyte interface. Thus, R ct is a charge transfer resistance and C dl represents the interfacial (double layer) capacitance. The impedance parameters R f and C f correspond to the resistance and capacitance of the inner part of the barrier composite film. C dl of TiO 2 -SnS 2 -HT is several times larger when DCF is present in the electrolyte, indicating a strong and large adsorption extent on the film's surface. The changes in C f and R f values are ascribed to different dielectric properties upon formation of the thin adsorption layer. On the other hand, the presence of DCF molecules in the electrolyte influenced R Ω and R ct values of TiO 2 -SnS 2 -COMM film-electrolyte interface as can be seen in high frequency part of EIS magnitude plots ( Figure 3). The impact on the film's inner part, according to R f and C f values was negligible also confirming the less degree of DCF adsorption on TiO 2 -SnS 2 -COMM film.
large adsorption extent on the film's surface. The changes in Cf and Rf values are ascribed to different dielectric properties upon formation of the thin adsorption layer. On the other hand, the presence of DCF molecules in the electrolyte influenced RΩ and Rct values of TiO2-SnS2-COMM film-electrolyte interface as can be seen in high frequency part of EIS magnitude plots (Figure 3). The impact on the film's inner part, according to Rf and Cf values was negligible also confirming the less degree of DCF adsorption on TiO2-SnS2-COMM film.   Table 4. Impedance parameters of TiO 2 -SnS 2 catalyst films recorded in NaCl and NaCl + DCF solution (R Ω , C dl , R ct , C f , and R f stand for ohmic resistance, double layer capacitance, charge transfer resistance, capacitance and resistance of the composite film, respectively)

Solar-Driven Photocatalytic Treatment of DCF Using TiO 2 -SnS 2 Catalysts
The photoactivity of TiO 2 -SnS 2 -COMM and TiO 2 -SnS 2 -HT composites under simulated solar irradiation was studied using DCF as a model organic pollutant. The influence of key process parameters; initial pH, composite formulation and concentration of H 2 O 2 oxidant, on the DCF removal/conversion yield was evaluated. The yield results were compared to that obtained using benchmark material, TiO 2 P25. The determined kinetic profiles of DCF removal by solar/TiO 2 , solar/TiO 2 -SnS 2 -COMM, solar/TiO 2 -SnS 2 -COMM/H 2 O 2 , solar/TiO 2 -SnS 2 -HT, and solar/TiO 2 -SnS 2 -HT/H 2 O 2 processes are given in Figure 4 and Figures S3-S6 (Supplementary Material), respectively. Two common effects can be observed; (i) the DCF adsorption is closely related to initial pH and (ii) kinetic profiles of DCF removal upon exposure to solar irradiation obeyed first-order kinetics regardless the catalyst material type and formulation, as well as conditions applied. The latter is in good agreement with the literature [40]. The photoactivity of TiO2-SnS2-COMM and TiO2-SnS2-HT composites under simulated solar irradiation was studied using DCF as a model organic pollutant. The influence of key process parameters; initial pH, composite formulation and concentration of H2O2 oxidant, on the DCF removal/conversion yield was evaluated. The yield results were compared to that obtained using benchmark material, TiO2 P25. The determined kinetic profiles of DCF removal by solar/TiO2, solar/TiO2-SnS2-COMM, solar/TiO2-SnS2-COMM/H2O2, solar/TiO2-SnS2-HT, and solar/TiO2-SnS2-HT/H2O2 processes are given in Figures 4 and S3-S6 (Supplementary Material), respectively. Two common effects can be observed; (i) the DCF adsorption is closely related to initial pH and (ii) kinetic profiles of DCF removal upon exposure to solar irradiation obeyed first-order kinetics regardless the catalyst material type and formulation, as well as conditions applied. The latter is in good agreement with the literature [40].   (Tables S1-S4, Supplementary Material). The corresponding graphs presenting recorded DCF conversions during the same treatments, performed at the same conditions are given in Figure S7 (Supplementary Material). As can be seen, the adsorption affinity of applied materials, including TiO 2 P25 (Figure 4), toward DCF in a dark strongly increases by lowering initial pH values. Such effect is related to the surface charge and specific surface area of material, as well as the dissociation constant of DCF. TiO 2 -SnS 2 -COMM has pH PZC value of 6.31, while pH PZC of TiO 2 -SnS 2 -HT is lower (4.61) ( Table 3). Furthermore, TiO 2 -SnS 2 -HT has almost double larger specific surface area than TiO 2 -SnS 2 -COMM (Table 2). EIS measurements clearly demonstrated that C dl of TiO 2 -SnS 2 -HT is several times larger when DCF is present in the electrolyte (230.6 >> 61.5 µF cm −2 , Table 4), indicating a strong and large adsorption extent on the film's surface, which was not the case with TiO 2 -SnS 2 -COMM (63.8 > 45.0 µF cm −2 ). Trovo and Nogueira [41] stated that pKa of DCF is 4.15, meaning that at pH > pKa DCF would be present mostly in its deprotonated form and vice versa. These facts mean that the adsorption of DCF (in its deprotonated form) would be promoted at pH < pH PZC of studied materials. The strongest adsorptions in our case, i.e., DCF removal extents during the initial dark period, were obtained at the lowest studied pH (4) (Figure 5), especially in the case of TiO 2 -SnS 2 -COMM. This material exhibited rather limited adsorption capacity of DCF at other studied pH values (5.5 and 7, Figure 5). Similar results are obtained by benchmark material TiO 2 P25 (Figure 4), which is built into TiO 2 -SnS 2 -COMM. On the other hand, TiO 2 -SnS 2 -HT showed rather high affinity toward DCF adsorption during initial dark period at pH 5.5 as well ( Figure 5). Such results are the consequence of abovementioned differences in BET surface area ( Table 2), but also the permeability of immobilized films, as demonstrated by EIS measurements (Figure 3, Table 4). It should be also noted that we have studied the initial pH of DCF solution, thus the immersion of glass plate with immobilized TiO 2 -SnS 2 composite resulted with the certain decrease of adjusted initial pH toward more acidic pH values (e.g., from 5.5 to 4.65 in the case of TiO 2 -SnS 2 -HT). As can be seen from the results showed in Figure 5, and DCF removal by TiO 2 P25 (Figure 4), the significant overall DCF removal (≥70%) was accomplished when rather large portion of DCF (≥40%) was adsorbed during initial dark period. Such an effect is particularly pronounced in the case of processes using TiO 2 -SnS 2 -COMM composite. Usually such high removals were followed with rather high DCF conversions as well (Figure 4 and Figure S7, Supplementary Material). The importance of adsorption during dark period and its relationship with high overall removal/conversion upon exposure to solar irradiation can be related with the mechanism of reactive species generation within photocatalytic process. Upon illumination of semiconducting material with the sufficient high energy, hν ≥ E g , electron and hole pairs (e − /h + ) formed at catalyst surface enable the generation of radical species or direct oxidation/reduction of adsorbed organics [42]. Hence, effective adsorption of organics during dark period enables their direct degradation at surface either at e − /h + or by formed radicals, overcoming transfer limitations related to their diffusion into the solution where bulk organics can react with them [40,43]. Improved photocatalytic activity of TiO 2 -SnS 2 -HT composite over TiO 2 -SnS 2 -COMM can be attributed to (i) higher number of free charge carriers in TiO 2 -SnS 2 -HT composite ( Figure 2) and (ii) lower R ct values pointing to facilitated charge transfer across the film (Table 4). A decrease in R ct value, when DCF is present, could indicate charge transfer between TiO 2 and SnS 2 hetero-junctions mediated by DCF. It is also noteworthy to stress that both composite films possess the excellent corrosion resistance (R corr ), which is a key factor for their application as catalysts in photocatalytic processes for water purification purposes. R corr equals to the sum of the resistance components R ct and R f ( Table 4) and for both film is in order of MΩ cm 2 . This result also displays the durability of TiO 2 -SnS 2 interaction to the inhibition of SnS 2 leaching and in turn materials' (catalysts') photostability. In conclusion, EIS as non-destructive in situ method clearly showed that TiO 2 -SnS 2 -HT composite film exhibits high DCF adsorption degree and offers good electric and charge transfer properties which are key-governing factor for the enhanced efficient photocatalytic activity in the solar driven water treatment. Comparing the effectiveness of processes using composites ( Figure 5 and Figure S7, Supplementary Material) to that with benchmark material (Figure 4.), following can be concluded. The adsorption of DCF during initial dark period was similar at the lowest initial pH studied (~50% of DCF was removed), regardless the material used. The increase in SnS 2 wt % yielded an increase in overall DCF removal and conversion comparing to process using benchmark material, presumably due to higher photoactivity of composites under induced simulated solar irradiation.
In order to provide deeper insight into effects occurring in TiO 2 -SnS 2 composites that are responsible for its activity within solar driven water treatment, the semiconducting properties of pure components TiO 2 -HT and SnS 2 -HT were also investigated using Mott-Schottky analysis ( Figure 6A). Both catalyst films exhibit positive slope of C −2 SC against E dependence, suggesting n-type semiconducting behavior. From the slope and the intercepts values of Mott-Schottky plots ( Figure 6A In order to provide deeper insight into effects occurring in TiO2-SnS2 composites that are responsible for its activity within solar driven water treatment, the semiconducting properties of pure components TiO2-HT and SnS2-HT were also investigated using Mott-Schottky analysis ( Figure 6A). Both catalyst films exhibit positive slope of against E dependence, suggesting n-type semiconducting behavior. From the slope and the intercepts values of Mott-Schottky plots ( Figure  6A), ND and EFB were calculated using Equation (1): ND values are determined as 1.6 × 10 18 cm −3 and 2.6 × 10 20 cm −3 , and EFB as −0.61 V and −0.77 for TiO2 and SnS2, respectively.  Taking into account the semiconducting parameters; E FB values and E g values (Table 2), the energy band diagram was constructed ( Figure 6B). The conduction band (E C ) and valence band (E V ) potentials of SnS 2 are more negative than those of TiO 2 . This thermodynamically allows the photogenerated electrons to transfer from the SnS 2 conduction band to TiO 2 conduction band under visible-light irradiation (λ > 420 nm) that enhances the separation of photogenerated electron and holes in SnS 2 and bring about the sensitization of TiO 2 . Additionally, the positions of the decomposition potentials of the semiconductor in competition with other redox reactions in the solution were calculated from the thermodynamic data [44][45][46]. Taking into account the semiconducting parameters; EFB values and Eg values (Table 2), the energy band diagram was constructed ( Figure 6B). The conduction band (EC) and valence band (EV) potentials of SnS2 are more negative than those of TiO2. This thermodynamically allows the photogenerated electrons to transfer from the SnS2 conduction band to TiO2 conduction band under visible-light irradiation (λ > 420 nm) that enhances the separation of photogenerated electron and holes in SnS2 and bring about the sensitization of TiO2. Additionally, the positions of the decomposition potentials of the semiconductor in competition with other redox reactions in the solution were calculated from the thermodynamic data [44][45][46].
According to Gerischer's approach developed for the binary compound semiconductors [38,47], the reductive decomposition potentials (nEdecomp) of both sulfide film and oxide film lie outside of the band edges, hence both semiconductor components are stable towards theirs reductive decomposition processes, which take place by electrons as major charge carriers in n-type semiconductor 2 2 According to Gerischer's approach developed for the binary compound semiconductors [38,47], the reductive decomposition potentials ( n E decomp ) of both sulfide film and oxide film lie outside of the band edges, hence both semiconductor components are stable towards theirs reductive decomposition processes, which take place by electrons as major charge carriers in n-type semiconductor At first glance, the positions of the anodic decomposition potential values ( p E decomp ) may suggest the susceptibility of both semiconductor materials to the anodic decomposition; however, when discussing the materials' stability all competitive redox process at the film-electrolyte interface must be taken into consideration. If redox potentials of solvent are located above p E decomp and if these reactions are fast enough, decomposition of the semiconductor would be stabilized in the dark and during solar irradiation conditions [47]. As can be seen from the energy band diagram constructed, based on the redox potential (Fermi level position), the most thermodynamically favorable reaction is the formation of hydroxyl radical, HO• via reaction that is desirable for DCF conversion process. The position of calculated energy levels for the anodic decomposition potential ( p E decomp ) of SnS 2 and for reaction suggest these are competitive process, thus latter reaction is kinetically more favorable, i.e., formation of HO• is more favorable than SnS 2 oxidative decomposition. Likewise, the oxidative decomposition of TiO 2 TiO 2 + 4Cl − + 4h + VB → TiCl 4 (aq) + O 2 (9) and O 2 evolution reaction are also competitive processes with the latter being thermodynamically and kinetically favorable. In summary, it should be stressed out that even though if conditions of solar irradiation lead to SnS 2 oxidative decomposition, the composite catalyst material would remain stable and photo-catalytically active due to stable TiO 2 phase/component since p E decomp of TiO 2 in the presence of chloride ions is below p E decomp of SnS 2 . Additionally, reactions (7) and (9) take place using holes as minor charge carriers in the n-type semiconductor.
Since no clear pattern of behavior and influence of studied process parameters (initial pH, composite formulation, and presence and concentration of H 2 O 2 ) can be observed neither from the kinetic profiles ( Figures S3-S6, Supplementary Material) nor from the end treatment points (providing insight into pH influence only) ( Figures 5 and S7, Supplementary Material), the RSM approach was employed. The statistical part on derived RSM models is summarized below. RSM models (M1-M8) were derived (Table S6, Supplementary Material) by applying the multiple regression analysis on FFD or BBD matrices and DCF removal and conversion extents after 60 min treatment (Tables S1-S4, Supplementary Material). The models are characterized for their accuracy, significance, and predictivity by ANOVA (0.982 < R 2 < 0.998 and 0.0001 < p < 0.0054; Tables S7-S10, Supplementary Material) and RD tools (example provided in Figure S8, Supplementary Material). According to obtained results, models can be used hereinafter as tools to enlighten the influence of studied parameters of on DCF removal and conversion, correlating determined semiconducting and barrier properties of applied semiconducting materials and occurring mechanisms within the solar driven photocatalytic treatment.
The mutual effects of studied process parameters on DCF removal and conversion are presented through 3D surface and contour plots. As discussed above, the pH has significant influence on water treatment by studied processes, particularly on the adsorption during initial dark period and consequently on the overall yield. This effect can be observed from 3D surface and contour plots describing overall  (Figure 10). ANOVA indicated the same trend as well (Tables S7-S10, Supplementary Material). ANOVA also revealed that both H 2 O 2 (where applicable) and SnS 2 wt % are significant process parameters. The importance of SnS 2 wt % in composite can be related to the decrease in band gap values by increasing SnS 2 content within composite ( Table 2). The composites with the lowest SnS 2 content (5 wt %) exhibited similar effectiveness as TiO 2 P25 when treatments were conducted in the absence of H 2 O 2 (Figures 4, 7 and 9). DCF removal and conversion extents obtained at pH 4 were~70% and~60%, respectively. However, the increase in SnS 2 content increased effectiveness of process using TiO 2 -SnS 2 -COMM almost linearly reaching 81% and 67% of DCF removal and conversion, respectively, at pH 4 and 50 wt % of SnS 2 (Figure 7). TiO 2 -SnS 2 -HT showed even higher effectiveness at comparable conditions (93% and 77% of DCF removal and conversion, respectively) ( Figure 9). Its effectiveness increase follows the exponential trend, similar as noticed in the decrease of band gaps values ( Table 2). The addition of H 2 O 2 into the system improved its performance for approximately 10% in both DCF removal and conversion, regardless the composite preparation method (Figures 7-10). According to ANOVA (Tables S8 and S10, Supplementary Material) [H 2 O 2 ] is a significant process parameter, while the constructed energy-band diagram ( Figure 6B) clearly shows favorability of HO• generation through Equations (6) and (8). Due to the fact that no changes in H 2 O 2 during the treatment were observed (results not showed), one can assume that either (i) H 2 O 2 was not consumed or (ii) H 2 O 2 was simultaneously consumed and produced within the system. In the former case, the observed positive effect of H 2 O 2 (Figures 7-10) can be explained by playing a role as a weak acid, thus lowering pH and consequently promoting favorable conditions for the adsorption during the initial dark period. In turn, subsequent conversion at the catalyst surface during the irradiation period is improved as well. However, the latter assumption on simultaneous consumption/generation is more likely to occur as it is supported by the energy band diagram ( Figure 6B). The generation of HO• through reactions involving H 2 O 2 depletion by photogenerated electrons are the most thermodynamically and kinetically favorable anodic reactions ( Figure 6B). The optimal wt % of SnS 2 is also influenced by H 2 O 2 . A lower wt % of SnS 2 is required to achieve maximum DCF removal and conversion in the presence of H 2 O 2 . In the case of solar/TiO 2 -SnS 2 -HT/H 2 O 2 36.1 wt % SnS 2 was found to be optimal, while 46.5 wt % was deemed optimal for solar/TiO 2 -SnS 2 -HT, according to Figures 9 and 10, respectively. The plausible explanation for such effect can be found in the competitive reactions occurring at anode (7) and (8), as shown through constructed energy band diagram ( Figure 6B). Hence, when oxidant is absent, more SnS 2 would be exposed to decomposition by photogenerated h + (7), thus this negative effect should be compensated with higher SnS 2 content within the composite. On the other hand, SnS 2 decomposition is suppressed in the presence of H 2 O 2 due its favorable depletion by photogenerated e − (8).
describing overall DCF removal and conversion by solar/TiO2-SnS2-COMM (Figure 7), solar/TiO2-SnS2-COMM/H2O2 (Figure 8), solar/TiO2-SnS2-HT (Figure 9), and solar/TiO2-SnS2-HT/H2O2 ( Figure  10). ANOVA indicated the same trend as well (Tables S7-S10, Supplementary Material). ANOVA also revealed that both H2O2 (where applicable) and SnS2 wt % are significant process parameters. The importance of SnS2 wt % in composite can be related to the decrease in band gap values by increasing SnS2 content within composite ( Table 2). The composites with the lowest SnS2 content (5 wt %) exhibited similar effectiveness as TiO2 P25 when treatments were conducted in the absence of H2O2 (Figures 4, 7 and 9). DCF removal and conversion extents obtained at pH 4 were ~70% and ~60%, respectively. However, the increase in SnS2 content increased effectiveness of process using TiO2-SnS2-COMM almost linearly reaching 81% and 67% of DCF removal and conversion, respectively, at pH 4 and 50 wt % of SnS2 (Figure 7). TiO2-SnS2-HT showed even higher effectiveness at comparable conditions (93% and 77% of DCF removal and conversion, respectively) ( Figure 9). Its effectiveness increase follows the exponential trend, similar as noticed in the decrease of band gaps values ( Table 2). The addition of H2O2 into the system improved its performance for approximately 10% in both DCF removal and conversion, regardless the composite preparation method (Figures 7-10). According to ANOVA (Tables S8 and S10, Supplementary Material) [H2O2] is a significant process parameter, while the constructed energy-band diagram ( Figure 6B) clearly shows favorability of HO• generation through Equations (6) and (8). Due to the fact that no changes in H2O2 during the treatment were observed (results not showed), one can assume that either (i) H2O2 was not consumed or (ii) H2O2 was simultaneously consumed and produced within the system. In the former case, the observed positive effect of H2O2 (Figures 7-10) can be explained by playing a role as a weak acid, thus lowering pH and consequently promoting favorable conditions for the adsorption during the initial dark period. In turn, subsequent conversion at the catalyst surface during the irradiation period is improved as well. However, the latter assumption on simultaneous consumption/generation is more likely to occur as it is supported by the energy band diagram ( Figure 6B). The generation of HO• through reactions involving H2O2 depletion by photogenerated electrons are the most thermodynamically and kinetically favorable anodic reactions ( Figure 6B). The optimal wt % of SnS2 is also influenced by H2O2. A lower wt % of SnS2 is required to achieve maximum DCF removal and conversion in the presence of H2O2. In the case of solar/TiO2-SnS2-HT/H2O2 36.1 wt % SnS2 was found to be optimal, while 46.5 wt % was deemed optimal for solar/TiO2-SnS2-HT, according to Figures 9 and 10, respectively. The plausible explanation for such effect can be found in the competitive reactions occurring at anode (7) and (8), as shown through constructed energy band diagram ( Figure 6B). Hence, when oxidant is absent, more SnS2 would be exposed to decomposition by photogenerated h + (7), thus this negative effect should be compensated with higher SnS2 content within the composite. On the other hand, SnS2 decomposition is suppressed in the presence of H2O2 due its favorable depletion by photogenerated e − (8).

Conclusions
The TiO2-SnS2 composite catalysts were prepared by hydrothermal synthesis (TiO2-SnS2-HT) and by immobilization of the commercial TiO2 and SnS2 particles (TiO2-SnS2-COMM) and applied for the solar-driven DCF degradation.
The first property relevant to the photocatalytic activity of TiO2-SnS2 photocatalyst is its energy band configuration. Energy band diagram was constructed using semiconducting parameters (results of DRS and Mott-Schottky analysis) as well as the thermodynamic data for the determination of the energy levels for the catalysts' oxidative and reductive decomposition and competitive redox reactions within the band gap. Based on these results, the catalysts' photocorrosion stability and kinetically preferable redox reactions were discussed in the terms of the solar-driven DCF degradation.
The difference in the conduction band edges of the SnS2 and TiO2 composite components of approx. 160 mV enabled the transfer of photogenerated electrons from the conduction band of the sensitized narrow-band gap semiconductor SnS2 to the conduction band of the wide-band gap TiO2 semiconductor. Thus, SnS2-TiO2 photocatalysts with the heterojunction structure satisfy the light adsorption and charge separation criteria simultaneously, resulting in an efficient photoactivity.
Another key issue influencing the photocatalytic capability of SnS2-TiO2 photocatalyst is a nature of the surface-interface chemistry. The increased values of the specific area surface (BET analysis) and the adsorption capacity at the solid-solution interfaces (EIS results) and the lower pHpzc values (potentiometric titration) revealed that composite catalyst prepared by HT synthesis shows remarkable improvement in the effectiveness of the DCF removal and conversion during solar-driven water treatment. More SnS2-TiO2 heterojunctions facilitated the effective photoelectron transfer, which promoted photocatalytic DCF degradation.
Furthermore, thermodynamically favored reactions in the energy band diagrams elucidated the observed effects during the photocatalytic DCF treatment. In the presence of H2O2 less wt % SnS2 in the composite material is required to achieve the highest DCF removal and conversion degree, which is contributed to competitive depletion of H2O2 by the photogenerated electron. Although the effectiveness of TiO2-SnS2-COMM is fairly limited at circumneutral conditions, both composites are

Conclusions
The TiO 2 -SnS 2 composite catalysts were prepared by hydrothermal synthesis (TiO 2 -SnS 2 -HT) and by immobilization of the commercial TiO 2 and SnS 2 particles (TiO 2 -SnS 2 -COMM) and applied for the solar-driven DCF degradation.
The first property relevant to the photocatalytic activity of TiO 2 -SnS 2 photocatalyst is its energy band configuration. Energy band diagram was constructed using semiconducting parameters (results of DRS and Mott-Schottky analysis) as well as the thermodynamic data for the determination of the energy levels for the catalysts' oxidative and reductive decomposition and competitive redox reactions within the band gap. Based on these results, the catalysts' photocorrosion stability and kinetically preferable redox reactions were discussed in the terms of the solar-driven DCF degradation.
The difference in the conduction band edges of the SnS 2 and TiO 2 composite components of approx. 160 mV enabled the transfer of photogenerated electrons from the conduction band of the sensitized narrow-band gap semiconductor SnS 2 to the conduction band of the wide-band gap TiO 2 semiconductor. Thus, SnS 2 -TiO 2 photocatalysts with the heterojunction structure satisfy the light adsorption and charge separation criteria simultaneously, resulting in an efficient photoactivity.
Another key issue influencing the photocatalytic capability of SnS 2 -TiO 2 photocatalyst is a nature of the surface-interface chemistry. The increased values of the specific area surface (BET analysis) and the adsorption capacity at the solid-solution interfaces (EIS results) and the lower pHpzc values (potentiometric titration) revealed that composite catalyst prepared by HT synthesis shows remarkable improvement in the effectiveness of the DCF removal and conversion during solar-driven water treatment. More SnS 2 -TiO 2 heterojunctions facilitated the effective photoelectron transfer, which promoted photocatalytic DCF degradation.
Furthermore, thermodynamically favored reactions in the energy band diagrams elucidated the observed effects during the photocatalytic DCF treatment. In the presence of H 2 O 2 less wt % SnS 2 in the composite material is required to achieve the highest DCF removal and conversion degree, which is contributed to competitive depletion of H 2 O 2 by the photogenerated electron. Although the effectiveness of TiO 2 -SnS 2 -COMM is fairly limited at circumneutral conditions, both composites are effective for DCF removal and conversion at acidic pHs and showed higher activity at comparable conditions under solar irradiation than benchmark material TiO 2 P25. This is particularly valid for TiO 2 -SnS 2 -HT which was more active than TiO 2 P25 through the entire range of parameters studied.