Exploring Synthesis Methods of CdS/TiO₂ Photocatalysts for Enhanced Hydrogen Production Under Visible Light

Abstract

TiO₂ was synthesized via the sol–gel method and employed as a support material for the deposition of CdS nanofibers using two novel techniques: impregnation and photodeposition. XRD characterization shows that crystallite size decreases when CdS is incorporated into TiO₂. UV-Vis spectroscopy showed that the bandgap of the CdS/TiO₂ heterostructured nanocomposites decreases compared to the raw TiO₂ support, making them very appropriate for photocatalytic applications in the visible region. The photocatalysts were tested for hydrogen production in methanol–water solutions under visible light conditions. It was observed that the TiC20 photocatalyst prepared by the impregnation method improved the photocatalytic activity compared with photodeposition technique (TiC20FD), achieving a maximum hydrogen production of 570.5 µmol H₂ ${\mathrm{g}}_{\mathrm{c}\mathrm{a}\mathrm{t}}^{-1}$ h⁻¹, while the latter attained 383.4 µmol H₂ ${\mathrm{g}}_{\mathrm{c}\mathrm{a}\mathrm{t}}^{-1}$ h⁻¹.

1. Introduction

Given the growing energy demand and environmental concerns, there is an increasing need to explore alternative and environmentally sustainable energy sources [1]. Among these alternatives, hydrogen has emerged as one of the most promising candidates to replace traditional fossil fuels, owing to its high energy efficiency and the fact that its combustion produces no greenhouse gases [2,3]. As a result, photocatalytic hydrogen production has garnered significant attention, as it leverages solar energy—a renewable and virtually limitless resource [4,5].

The development of semiconductor-based photocatalysts for hydrogen production has been extensively investigated. Among these, titanium dioxide (TiO₂) is widely recognized for its high stability, low cost, non-toxicity [6,7], and favorable conduction band, allowing it to produce hydrogen through water reduction [8]. However, its photocatalytic activity is restricted to the ultraviolet (UV) region, accounting for only ~5% of the solar spectrum [9], limiting their application in the visible range. Several attempts have been made to overcome its responsiveness and apply it in the visible light range; TiO₂ is often coupled with narrow-bandgap semiconductors, forming heterostructured compounds that broaden light absorption into the visible light range [10,11]. TiO₂-based heterostructures can be used in this region of electromagnetic spectrum. Representative examples include CdS [12,13], Fe₂O₃ [14], and WO₃ [15]. Among these, CdS is particularly promising due to its suitable bandgap (2.4 eV), which facilitates efficient visible-light absorption, and its sufficiently negative conduction band potential (−0.51 V vs. SHE) for proton (H⁺) reduction [16]. Consequently, TiO₂/CdS heterostructures have gained considerable attention for their enhanced photocatalytic efficiency [17].

The synthesis method of CdS/TiO₂ materials plays a critical role in determining their hydrogen production efficiency. For example, CdS/TiO₂ composites prepared via the solvothermal method exhibit mesoporous structures, with irregular spherical morphologies and high photocatalytic activity [18]. Likewise, the in situ synthesis of TiO₂ nanofibers, followed by CdS deposition using cadmium nitrate as a precursor, shows hydrogen production efficiency enhanced by a factor of four compared to pure CdS [19]. Furthermore, the construction of Z-scheme photocatalytic systems via the successive ionic layer adsorption and reaction (SILAR) method improves the charge separation efficiency of CdS/TiO₂, further enhancing its photocatalytic performance [20].

In this work, the incorporation of a low-cost semiconductor (CdS) onto TiO₂ is investigated by synthesizing CdS/TiO₂ composites using two novel methods—impregnation (TiCX) and photodeposition (TiCXFD)—to identify the most efficient approach for photocatalytic hydrogen production, where X denotes the CdS content in weight percent. The resulting materials were systematically characterized using X-ray diffraction (XRD), Fourier transform infrared spectroscopy (FTIR), energy-dispersive X-ray fluorescence (EDXRF), nitrogen physisorption, UV-Vis spectroscopy, scanning electron microscopy (SEM), photoluminescence (PL), (photo)electrochemical techniques, and X-ray photoelectron spectroscopy (XPS). Finally, their photocatalytic performance was assessed for hydrogen production under visible-light irradiation using a 50:50 %vol methanol–water solution as a sacrificial agent.

2. Results

2.1. X-Ray Diffraction (XRD)

Figure 1 presents the X-ray diffraction (XRD) spectra of TiO₂, TiC10, TiC20, TiC20FD, TiC30, and CdS, respectively. The diffraction patterns of TiO₂, TiC10, TiC20, TiC20FD, and TiC30 exhibit characteristic peaks corresponding to the anatase phase of TiO₂ (COD ID: 7206075) at 2θ = 25.4°, 37.9°, 38.7°, 48.1°, 54.0°, 55.2°, 62.8°, and 68.9°, which are associated with the (001), (004), (112), (200), (105), (211), (204), and (116) crystallographic planes, respectively. A minor brookite phase was identified at 2θ = 30.9°, indexed to the (121) crystallographic plane (COD ID: 9004137). Additionally, diffraction peaks were observed at 2θ = 26.7°, 28.4°, 36.9°, 44.0°, 48.2°, and 52.3°, corresponding to the (002), (101), (102), (110), (103), and (112) planes, respectively—characteristic of the hexagonal phase of CdS (COD ID: 1011054).

Table 1. Lattice parameters of TiO₂, CdS, TiC10, TiC20, TiC20FD, and TiC30 materials.

Sample	Lattice Parameters		Crystallite Size (Å) (001)
	CdS (Å)	TiO2 (Å)
TiO2	-	a = b = 3.78, c = 9.50	170
CdS	a = b = 4.10, c = 6.63	-	1352
TiC10	a = b = 4.10, c = 6.66	a = b = 3.78, c = 9.48	154
TiC20	a = b = 4.16, c = 6.67	a = b = 3.78, c = 9.49	148
TiC20FD	a = b = 4.11, c = 6.67	a = b = 3.78, c = 9.47	154
TiC30	a = b = 4.11, c = 6.68	a = b = 3.78, c = 9.49	125
TiC20 after Cycle 4	a = b = 4.15, c = 6.65	a = b = 3.78, c = 9.48	150

Crystallite size was determined from the (001) plane corresponding to TiO₂ and for CdS from the (002) plane.

presents the lattice parameters of TiO₂, with “a” calculated from the (200) plane and “c” from the (004) plane, and CdS, with “a” calculated from the (002) plane and “c” from the (101) plane, along with the crystallite size of the synthesized materials, determined using the Scherrer equation. A comparison of the a, b, and c parameters of CdS and TiO₂ in the TiC10, TiC20, TiC20FD, and TiC30 samples with those of pure CdS and TiO₂ indicates no significant changes, suggesting that Cd²⁺ ions are not incorporated into the TiO₂ lattice. Consequently, it can be inferred that CdS is primarily deposited on the TiO₂ surface. Additionally, XRD analysis was performed for the TiC20 material after 4 reaction cycles to corroborate its crystalline structure.

2.2. Chemical Analysis

Table 2. Weight content of Cd, S and TiO₂ in the TiC10, TiC20, TiC20FD, and TiC30 materials, and their corresponding specific surface areas and bandgap widths.

Sample	Cd (wt.%)	S (wt.%)	TiO2 (wt.%)	Eg (eV)	SBET (m2/g)
TiO2	-	-	100.0	3.32	51
TiC10	8.4	2.4	89.2	2.53	72
TiC20	17.9	4.9	77.2	2.50	83
TiC20FD	18.3	4.0	77.3	2.53	87
TiC30	26.5	7.1	66.4	2.49	88
CdS	21.5	78.5	-	2.52	152

presents the CdS and TiO₂ weight percentages contents of the TiC10, TiC20, TiC30, and TiC20FD materials, with values closely aligned to the theoretically expected nominal ones. This suggests that the synthesis methods (impregnation and photodeposition) successfully facilitate the effective deposition of CdS onto TiO₂.

2.3. Diffuse Reflectance Spectroscopy (DRS)

Figure 2A presents the UV-Vis diffuse reflectance spectra of the TiO₂, TiC10, TiC20, TiC20FD, TiC30, and CdS materials. For the TiO₂ sample, an absorption edge is observed near 380 nm, while for CdS, the absorption edge is in the visible region (~500 nm). When comparing pure TiO₂ with the TiC10, TiC20, TiC20FD, and TiC30 samples, the strong absorption of visible light around 500 nm is observed, suggesting that the light absorption capability of CdS is retained. However, the materials synthesized via the impregnation method show higher light absorption compared to the TiC20FD material, which was synthesized by photodeposition. The enlargement of the TiO₂ spectrum in the 400–600 nm range (Figure 2B) reveals visible light absorption, which can be attributed to carbon species originating from the n-butanol solvent used during synthesis. It has been reported that the presence of carbon in TiO₂ can shift its absorption edge into the visible region [21].

The bandgap energy of the samples was determined using the Tauc plot and subsequently corrected using the Kubelka–Munk function (Equation (1)) [22]. In this equation, α represents the extinction coefficient, proportional to F(R) h is Planck’s constant; v is the frequency of electromagnetic radiation (s⁻¹); B is the absorption constant; and $E_g$ is the bandgap energy (eV). Indirect allowed transitions were considered for these materials, with n = 2.

α (hv) ≈ B (h − E_g)ⁿ

(1)

The bandgap energies of the materials were determined using the Tauc plot, with the absorption edge extrapolated to the x-axis. As depicted in Figure 2C, the TiO₂ and CdS samples exhibit bandgap energies of 3.32 eV and 2.52 eV, respectively, in close agreement with values reported in the literature [23,24]. The TiC10, TiC20, TiC20FD, and TiC30 samples display bandgaps of 2.53 eV, 2.50 eV, 2.53 eV, and 2.49 eV, respectively, suggesting the effective coupling of CdS onto TiO₂.

2.4. Textural Properties

The nitrogen adsorption–desorption isotherms for the TiO₂, TiC10, TiC20, TiC20FD, TiC30, and CdS materials are presented in Figure 3. All materials exhibit a type IV isotherm, indicative of mesoporosity. TiO₂ displays an H1-type hysteresis loop, which is typical of porous materials consisting of particles with cylindrical channels or aggregates of spheroidal particles with uniform pore size and shape [22]. In contrast, the CdS sample exhibits an H3-type hysteresis loop, characteristic of non-rigid aggregates of plate-like particles with non-uniform shapes [25]. The hysteresis loops for TiC10 and TiC20 are also of the H1 type, like TiO₂, suggesting that the impregnation method leads to better dispersion of CdS on TiO₂. The TiC30 material, however, presents an H3-type hysteresis loop, likely due to the higher content of CdS on TiO₂. The TiC20FD material shows an H3-type hysteresis loop as well, which suggests that the photodeposition method may result in poor dispersion of CdS on TiO₂, causing it to form clusters. This observation highlights that the morphology of the materials varies depending on the synthesis method employed. The specific surface areas of the materials were determined using the BET method, and the results are summarized in Table 2. CdS exhibits the highest surface area, attributed to the interstitial spaces between the CdS nanofibers. The TiC10, TiC20, TiC20FD, and TiC30 materials exhibit similar surface areas, but compared to TiO₂, they show an increase of approximately 21 to 36 m²/g, which can be attributed to the presence of CdS.

Table 2 summarizes the bandgap, surface area and chemical composition obtained by XRF.

2.5. Fourier Transform Infrared Spectroscopy (FTIR)

The FTIR spectra of the synthesized materials (TiO₂, TiC10, TiC20, TiC20FD, TiC30, and CdS) are presented in Figure 4. All spectra exhibit a band at 3437 cm⁻¹, which corresponds to the stretching vibrations of O-H groups [26]. These groups are present on the surface of the materials, originating from the n-butanol and/or water used in the synthesis process. The TiO₂, TiC10, TiC20, TiC20FD, and TiC30 materials show a band at 725 cm⁻¹, attributed to the stretching vibrations of Ti-O and Ti-O-Ti bonds [27,28]. In contrast, the pure CdS spectrum displays a band at 730 cm⁻¹, indicative of the Cd-S stretching vibration [29]. Additionally, all spectra of the synthesized materials reveal signals around 1035, 1341, and 1563 cm⁻¹, which are associated with the stretching vibrations of –C-N– and –NH₂ groups [30], respectively, which is certainly due to the ethylenediamine used as a template in the synthesis of CdS [31].

2.6. Photoluminescence (PL)

Figure 5 presents the photoluminescence spectra of the TiO₂, TiC10, TiC20, TiC20FD, TiC30, and CdS materials, obtained at an excitation wavelength of 299 nm. The TiO₂ sample displays a strong photoluminescent emission near 470 nm, indicative of significant electron–hole pair recombination, which can be attributed to various intrinsic lattice defects, such as oxygen vacancies, titanium vacancies, and interstitial defects [32,33,34]. For the CdS, TiC10, TiC20, TiC20FD, and TiC30 materials, an emission peak around 530 nm is observed, corresponding to the recombination of electrons trapped in sulfur vacancies with holes in the valence band of CdS [35]. The emission band centered about at 510 nm can be attributed to the surface trap induced emission, which involves the recombination of electrons trapped inside a sulfur vacancy with a hole in the valence band of the CdS nanofibers [36].

The emission peaks observed at 580 nm and 830 nm in CdS are typically attributed to the surface-trapped electron–hole pairs and/or recombination via surface defects in the CdS structure [37]. The band near 677 nm in CdS can be associated with surface-state-induced recombination and sulfur vacancies [38]. For TiO₂, the emission near 580 nm is likely related to surface oxygen vacancies [39,40], while the band around 677 nm can be ascribed to surface state emissions resulting from the recombination of electron–hole pairs at dangling bonds on TiO₂ nanoparticles [41]. The emission centered at 830 nm in TiO₂ is characteristic of the brookite phase—present only in trace amounts—and is associated with intrinsic lattice defects that act as trap sites [42]. The TiC20 and TiC20FD samples exhibit lower photoluminescence intensities compared to the pure CdS spectrum, suggesting that the photogenerated electrons in CdS are efficiently transferred to TiO₂, thereby reducing the recombination of the electron–hole pairs. However, the TiC20FD material shows a higher photoluminescence intensity than TiC20, which is likely due to the non-homogeneous distribution of CdS nanofibers resulting from the photodeposition method used to deposit CdS onto TiO₂ (TiC20FD), a consequence of the higher electron–hole pair recombination in TiC20FD than in TIC20.

2.7. XPS Analysis (X-Ray Photoelectron Spectroscopy)

X-ray photoelectron spectroscopy (XPS) was conducted to determine the oxidation states of the elements in the TiC20 material. Figure 6A presents the full XPS spectrum, revealing the presence of titanium, oxygen, cadmium, sulfur, and carbon. In Figure 6B, the deconvolution of the Ti 2p spectrum shows binding energies of 463.9 eV and 458.8 eV, corresponding to Ti 2p₁/₂ and Ti 2p₃/₂, which are characteristic of the Ti⁴⁺ oxidation state in TiO₂ [43]. The oxygen 1s spectra, shown in Figure 6C, display a peak at 530.1 eV, corresponding to the Ti–O bond, and an additional peak at 531.8 eV, indicating the presence of hydroxyl groups on the surface [44]. The binding energy peaks at 411.9 eV and 405.1 eV (Figure 6D) correspond to Cd 3d₃/₂ and Cd 3d₅/₂, respectively, confirming the presence of Cd²⁺ [45]. Likewise, the sulfur XPS spectra show peaks at 162.2 eV and 161.4 eV (Figure 6E), attributed to S 2p₁/₂ and S 2p₃/₂, respectively, confirming the presence of sulfide (S²⁻) [46]. The Ti and O spectra for TiO₂ are consistent with those obtained for the TiC20 material. Additionally, the spectrum for titanium, oxygen, cadmium, and sulfur in the TiCFD material exhibits binding energies identical to those observed in the TiC material, showing no significant differences.

2.8. Scanning Electron Microscopy (SEM), Energy-Dispersive X-Ray Spectroscopy (EDS), and Transmission Electron Microscopy (TEM)

Figure 7 presents the scanning electron microscopy (SEM) images. Figure 7A shows the TiO₂ after calcination with the formation of irregular agglomerates consisting of semi-spherical particles. In Figure 7B, the agglomeration of CdS nanofibers is shown; in Figure 7C, the composite TiC20 shows the presence of agglomerates of TiO₂ nanoparticles covered with the filamentous structures of CdS nanofibers.

The chemical composition of the TiC20 sample was analyzed using energy-dispersive X-ray spectroscopy (EDS). The elemental mapping images (Figure 8B–E) further demonstrate the uniform distribution of Cd and S on TiO₂, confirming the successful impregnation of CdS onto the TiO₂ surface.

The EDS spectrum of the TiC20, shown in Figure 8F, reveals the characteristic peaks corresponding to Ti, O, S, and Cd.

Table 3. Elemental analysis by EDS of TiC20.

Sample	Content	Elements
		Cd	S	Ti	O
TiC20	Wt.%	17.47	3.88	46.78	31.87

exhibits the quantitative microanalysis results, indicating a CdS content that closely matches the theoretical value. These findings are consistent with the results obtained from elemental analysis via X-ray fluorescence (XRF) (See Table 2), thereby confirming the presence of CdS on TiO₂.

Figure 9A shows the HRTEM image of the most active sample TiC20; one can observe the TiO₂ nanoparticle morphology, surrounded by the CdS fibers (Figure 9A). Figure 9B,C show the interplanar distances, measured using Gatan Microscopy Suite software version 3.60.4441.0. An interplanar spacing of 3.60 Å was observed, corresponding to the (101) plane of the anatase phase of TiO₂ [47], along with a spacing of 3.18 Å. This matches the (101) plane of the hexagonal close-packed phase of CdS [48].

2.9. (Photo)Electrochemical Characterization

Characterization was conducted under conditions comparable to those used for hydrogen production evaluation. Specifically, the same LED lamps (4 W, 450 nm) and a methanol–water solution (1:1, v/v) containing 0.1 M Na₂SO₄ were employed, with the materials deposited onto an FTO conductive substrate. To ensure that the responses observed in the (photo)electrochemical characterizations were primarily attributable to the analyzed materials, the FTO substrate was also characterized. Figure 10 presents the open-circuit potential (OCP) behavior of the materials under dark and illuminated conditions. The FTO substrate and TiO₂ exhibit minimal photoactivity upon illumination (Figure 10A) compared to CdS (Figure 10A) and the TiO₂ modified with this chalcogenide (Figure 10B). Additionally, illumination induces a shift in OCP toward more negative values, confirming that FTO, CdS, and TiO₂ are n-type semiconductors. For pure CdS, this negative shift diminishes progressively with each dark–light cycle, suggesting the increased recombination of photogenerated charge carriers. Otherwise, when CdS is coupled with TiO₂, recombination is significantly suppressed, as indicated by the nearly constant OCP shift across all dark–light cycles for the TiC10, TiC, TiC30, and TiCFD materials. This stability further highlights their excellent photoelectrochemical stability under the applied characterization conditions. Furthermore, the amount of CdS deposited on TiO₂ influences the photoresponse of the materials. For the TiC and TiCFD, the OCP shift under illumination is approximately twice that observed for the TiC10, whereas this shift decreases so that there are less negative values for the TiC30. These results suggest that CdS loading of 20 wt% on TiO₂ enhances the interaction between the two semiconductors, regardless of the deposition method used, as the behavior and OCP values under illumination are very similar for the samples obtained with this composition via impregnation (TiC) and photodeposition (TiCFD). To further elucidate the interaction between TiO₂ and different CdS loadings, the materials were analyzed using electrochemical impedance spectroscopy (EIS). Figure 11 presents the Nyquist plots obtained at OCP, where all materials exhibit high impedance under dark conditions. However, a significant decrease in impedance is exclusively observed upon illumination for chalcogenide and the samples containing it (Figure 11B). Additionally, the semicircles in the Nyquist plots for the TiC10 and TiC30 are larger than those for TiC and TiCFD, indicating higher charge transport and transfer resistance when the CdS content is either below or above 20 wt%. This trend can be attributed to the very low photoactivity of TiO₂ (Figure 10A), as its Nyquist spectra remain nearly unchanged under both dark and illuminated conditions, like those of the FTO substrate. This suggests that both materials exhibit inherently high charge transport and transfer resistance, irrespective of illumination. In contrast, when TiO₂ is coupled with CdS, these resistances decrease significantly. The Nyquist spectra of CdS show a drastic reduction in resistance upon illumination, a behavior associated with its strong photoresponse to visible LED light (Figure 11A). Thus, increasing the CdS content beyond 20 wt% results in a thicker CdS layer, which may partially obstruct light penetration in certain regions, thereby increasing charge transport and transfer resistance in these photocatalysts. On the other hand, the high resistance observed in the Nyquist plots, even under illumination, can be attributed to the use of Nafion as a binder to adhere the photocatalyst particles to the FTO substrate [49].

Mott–Schottky plots were obtained to estimate the energy band positions of the materials (Figure 12) and to analyze their interactions, as well as their ability to drive methanol oxidation and hydrogen production. Interestingly, the flat band potential of TiO₂ is nearly identical to that of the sample containing 10 wt% CdS, but shifts significantly to more negative values for the sample with 20 wt% CdS, before becoming less negative when the CdS content increases to 30 wt%. These trends suggest that TiO₂-CdS and CdS-CdS interactions are highly dependent on the CdS loading, directly affecting the band energy positions. To further illustrate this effect on the photocatalytic performance, a band alignment diagram was constructed (Figure 13). In this diagram, the conduction band position was approximated to the flat band potential, as all materials exhibited n-type semiconductor behavior (Figure 10). The valence band position was estimated by adding the bandgap value, determined from DRS (Figure 4), to the flat band potential (i.e. E_vb = E_fb + E_g). According to this diagram, all materials have the potential to drive methanol oxidation; however, only the conduction bands of CdS, TiC, and TiCFD are positioned at sufficiently favorable energy levels to enable hydrogen production.

2.10. Photocatalytic Evaluation

The photocatalytic activity of the TiO₂, TiC10, TiC20, TiC20FD, TiC30, and CdS materials is shown in Figure 14A. It is observed that TiO₂ does not exhibit photocatalytic activity, due to its bandgap of 3.32 eV, indicating that the material is not activated by the visible light energy used, as confirmed by the OCP characterization (Figure 10A). Pure CdS shows relatively low hydrogen production (73.9 µmol H₂ ${\mathrm{g}}_{\mathrm{c}\mathrm{a}\mathrm{t}}^{-1}$ h⁻¹), indicating limited photocatalytic activity. This behavior can be attributed to significant electron–hole recombination, as evidenced by the photoluminescence spectra (Figure 5) and the dark–light open-circuit potential (OCP) measurements (Figure 10A). In the photoluminescence spectra, pure CdS exhibited higher emission intensity compared to the TiC20 and TiC20FD samples, indicating a higher recombination rate. Moreover, in the OCP measurements, unlike the TiO₂/CdS composites, pure CdS was the only sample that did not fully recover its OCP value under illumination during each dark–light cycle, with the potential becoming progressively less negative in each cycle. In contrast, when CdS is coupled with TiO₂ via the impregnation method (TiC20), hydrogen production increases by a factor of 7.7, reaching a photoproduction rate of 570.5 µmol H₂ ${\mathrm{g}}_{\mathrm{c}\mathrm{a}\mathrm{t}}^{-1}$ h⁻¹. For the material synthesized by photodeposition (TiC20FD), hydrogen photoproduction improves by 5.2 times compared to pure CdS, yielding a photocatalytic hydrogen production activity of 383.4 µmol H₂ ${\mathrm{g}}_{\mathrm{c}\mathrm{a}\mathrm{t}}^{-1}$ h⁻¹ (Figure 14B). Nevertheless, the TiC20 material exhibits higher photocatalytic activity than TiC20FD, which can be attributed to the non-homogeneous distribution of CdS nanofibers on TiO₂ in the photodeposition process. Additionally, this material demonstrated a higher accumulation of photogenerated charge carriers and minimal recombination, along with one of the lowest charge transport and transfer resistances. Furthermore, its energy band alignment was particularly well-suited for facilitating methanol oxidation and hydrogen production. The low photocatalytic activity observed in TiC10 (14.2 µmol H₂ ${\mathrm{g}}_{\mathrm{c}\mathrm{a}\mathrm{t}}^{-1}$ h⁻¹) can be attributed to its low CdS content and a conduction band position that lacks sufficient energy to drive proton reduction to hydrogen. In contrast, the excess CdS in TiC30 increases charge transport and transfer resistances, while also modifying TiO₂-CdS and CdS-CdS interactions. This adjustment positions the conduction band at the threshold required for hydrogen generation, leading to inefficient charge transport and a decline in photocatalytic performance. As a result, TiC30 achieves a higher hydrogen production rate (94.1 µmol H₂ ${\mathrm{g}}_{\mathrm{c}\mathrm{a}\mathrm{t}}^{-1}$ h⁻¹) than TiC10 but remains less efficient than the TiC sample.

To compare the activity of our synthesized composites with similar CdS/TiO₂ systems,

Table 4. Summary of hydrogen production of synthesized materials compared to similar systems found in literature.

Material	Method	Sacrificial Agents	Light Source	Experimental Conditions	Hydrogen Production	Reference
1-CdS/TiO2	Hydrothermal	0.35 M Na2S/ 0.25 M Na2SO4	300 W Xe	10 mg/10 mL	1.07 mmol/gcath	[50]
2-CdS/TiO2	SILAR	Methanol	350 W Xe	50 mg/80 mL	1.02 mmol/gcath	[20]
3-CdS/TiO2	Hydrothermal	0.1 M Na2S/ 0.1 M Na2SO4	300 W Xe	10 mg/50 mL	1.494 mmol/gcath	[9]
4-CdS/TiO2	Two-step method	0.35 M Na2S/ 0.25 M Na2SO4	300 W Xe	50 mg/50 mL	1.048 mmol/gcath	[51]
5-CdS/TiO2	Hydrothermal	Lactic acid	300 W Xe	100 mg/100 mL	0.765 mmol/gcath	[52]
6-CdS/TiO2	SILAR	1.0 M Na2S/ 1.0 M Na2SO4	300 W Xe	80 mg/80 mL	0.678 mmol/gcath	[53]
7-TiC20FD	Photodeposition	Methanol	4 W	50 mg/200 mL	0.383 mmol/gcath	This work
8-TiC20	Impregnation	Methanol	4 W	50 mg/200 mL	0.570 mmol/gcath	This work

summarizes the photocatalytic hydrogen production of our composites and similar systems presented in some other scientific reports. It is worth mentioning that TiC20FD and TiC20 composites present competitive hydrogen productions under visible light and using much lower-energy lamps than the other reports, making this material attractive for hydrogen production using low power lamps, reducing the energy consumption.

2.11. Photocatalyst Stability

The stability of the TiC20 material was investigated, as it demonstrated the highest photocatalytic activity in hydrogen production. To assess its stability, four reaction cycles were conducted, each lasting 5 h, for a total time of 20 h, under identical reaction conditions, e.g., the same methanol–water ratio, and purging with N₂ for 15 min prior to each reaction. Figure 15A illustrates the evolution of the TiC20 after 20 h of reaction, while Figure 15B presents the hydrogen production in terms of µmol H₂ ${\mathrm{g}}_{\mathrm{c}\mathrm{a}\mathrm{t}}^{-1}$ h⁻¹. It was observed that after the second cycle, hydrogen production increased by 52.6% compared to the first cycle. For the third cycle, hydrogen production was similar to the first cycle, while in the fourth cycle, hydrogen production decreased by 25% relative to the first cycle. This behavior can be attributed to the oxidation of methanol (used as a sacrificial agent), which reacts with O₂ to form formaldehyde (Equation (2)). Formaldehyde has been reported as a more efficient sacrificial agent, significantly enhancing photocatalytic activity compared to methanol [54]. In the third and fourth cycles, formaldehyde is further oxidized to formic acid (HCOOH) (Equation (3)), which exhibits photocatalytic activity comparable to methanol. The reaction pathway culminates with complete formic acid decomposition (Equation (4)). The overall reaction mechanism for methanol decomposition follows the sequence described below [55]:

$$C{H}_3-O{H}_{\left(l\right)}\stackrel{hvcat}{\to }{H}_{2\left(g\right)}+HCO{H}_{\left(l\right)}$$

(2)

$$HCO{H}_{\left(l\right)}\stackrel{hvcat}{\to }{H}_{2\left(g\right)}+HCOO{H}_{\left(l\right)}$$

(3)

$$HCOO{H}_{\left(l\right)}\stackrel{hvcat}{\to }{H}_{2\left(g\right)}+C{O}_{2\left(g\right)}$$

(4)

The overall reaction is as follows:

$$C{H}_{3}O{H}_{\left(l\right)}+H_2{O}_{\left(l\right)}\stackrel{hvcat}{\to }{H}_{2\left(g\right)}+C{O}_{2\left(g\right)}$$

(5)

The results indicate the long-term stability and reusability of the material, as hydrogen production is still observed in the fourth reaction cycle. Furthermore, as shown in the XRD analysis (Figure 1), the material retains its crystalline structure without any detectable changes, confirming its structural stability.

2.12. Reaction Mechanism

The proposed mechanism for photocatalytic hydrogen production using TiC20 and TiC20FD materials is illustrated in Figure 16. The conduction band values for CdS and TiO₂ are presented according to values reported in the literature. Since the conduction band of CdS is more negative (BC ≈ −0.5 V/NHE) than that of TiO₂ (BC ≈ −0.24 V/NHE), the photogenerated electrons in the CdS conduction band are transferred to the TiO₂ conduction band, following the conventional charge transport mechanism in a heterojunction. This process enhances charge separation and minimizes recombination.

The TiC20 demonstrates higher efficiency because, during synthesis, Cd²⁺ is absorbed onto the active sites of TiO₂, remaining on its surface. Once absorbed, Cd²⁺ reacts with S²⁻ to form CdS. This process differs from conventional impregnation, as the resulting CdS has a greater likelihood of being in close contact with the active sites of TiO₂, where varying charge densities exist. Furthermore, CdS crystals grow directly on the active surface of TiO₂, enhancing interaction. In contrast, the TiC20FD material exhibits lower efficiency because the photodeposition of CdS onto TiO₂ (both materials being pre-formed and crystallized by different methods) results in a limited physical interaction between the phases, yielding only a modest synergistic charge transfer effect.

3. Materials and Methods

3.1. Synthesis of TiO₂

The TiO₂ photocatalyst was synthesized using the sol–gel method. In a flask, 18 mL of deionized water, 44 mL of n-butanol, and 0.2 mL of nitric acid (to adjust the pH to 3) were combined. Next, 44 mL of titanium (IV) butoxide from Merck (Darmstadt, Germany) was added dropwise to the solution. The mixture was then heated to 70 °C under reflux with continuous stirring for 24 h to allow the formation of the gel. The resulting xerogel was dried at 70 °C for 24 h. After drying, the solid was ground in an agate mortar to obtain a fine powder. Finally, the powder was calcined at 500 °C for 4 h at a heating rate of 2 °C/min.

3.2. Synthesis of CdS

CdS was prepared using the impregnation method. A 1:4 v/v mixture of water and n-butanol was prepared, to which 6.4 g of cadmium nitrate tetrahydrate was added. Ethylenediamine was then added dropwise in a 2:1 v/v ratio relative to n-butanol. After the dropwise addition, 1.25 mL of carbon disulfide was introduced. The resulting solution was heated to 100 °C for 1 h under magnetic stirring and subsequently cooled to room temperature while continuing the stirring process. The material was recovered by filtration, washed with a 50/50 v/v aqueous ethanol solution, and then dried at 80 °C for 1 h. All the reactives were aquired from Merck (Darmstadt, Germany).

3.3. Synthesis of CdS/TiO₂ Composites by Impregnation of Cd²⁺

The CdS/TiO₂ composites were synthesized via the impregnation method. TiO₂ was added to a 1:4 v/v mixture of water–n-butanol at room temperature under continuous stirring. An appropriate amount of cadmium nitrate tetrahydrate was subsequently introduced to achieve 10, 20, and 30 wt% CdS loading in TiO₂. Ethylenediamine was then added dropwise in a 2:1 v/v ratio relative to n-butanol. Following the dropwise addition, carbon disulfide was introduced in a 1:2 molar ratio with respect to Cd. The resulting solution was heated to 100 °C for 1 h under magnetic stirring and then allowed to cool to room temperature while stirring was maintained. The materials were recovered by filtration, washed with an aqueous ethanol solution (50/50 v/v), and finally dried at 80 °C for 1 h. The materials were labeled as TiC10, TiC20, and TiC30.

3.4. Synthesis of the CdS/TiO₂ Composite by Photodeposition

The synthesis was carried out to achieve a 20 wt% CdS loading in TiO₂. In a suitable vessel containing a methanol–water solution (1:4 v/v), 0.5 g of TiO₂ was added under nitrogen bubbling with continuous stirring. The pre-synthesized CdS was weighed and introduced into the solution. The suspension was then irradiated using a UV lamp (λ = 254 nm, 4.2 mW/cm²) for 4 h, while maintaining stirring and nitrogen flow. Finally, the material was recovered by centrifugation and dried at 80 °C for 1 h. The resulting material was labeled TiC20FD and was compared with TiC20, which showed better catalytic properties.

3.5. Characterization of the Materials

To determine the crystalline phases, the materials were analyzed using powder X-ray diffraction with a Bruker AXS D2 Phaser diffractometer (Billerica, MA, USA), equipped with a copper anode (Cu Kα = 1.5406 Å). The diffraction patterns were recorded in the 2θ range from 5° to 90° with a scan step size of 0.01° s⁻¹. The cadmium sulfide and titanium oxide contents were quantified using an X-ray fluorescence (XRF) spectrometer, JSX-1000S (JEOL, Akishima, Tokio, Japan), with Rh Kα radiation operating at 30 kV. Fourier transform infrared (FTIR) spectra of the samples were obtained using a Shimadzu IRAffinity-1 spectrometer (Kyoto, Japan), equipped with an attenuated total reflectance (ATR) module, with a resolution of 16 cm⁻¹ over the spectral range of 400–4000 cm⁻¹ and 200 scans. The specific surface areas were determined using the Brunauer–Emmett–Teller (BET) method by measuring nitrogen adsorption–desorption isotherms on a Quantachrome Autosorb-3B analyzer (Boynton Beach, FL, USA). Prior to adsorption, the samples were degassed in a vacuum (10⁻⁵ Torr) at 80 °C for 10 h. The optical properties of the materials were analyzed by diffuse reflectance UV-Vis spectroscopy in the range of 190–900 nm, using a Varian Cary 100 Scan spectrophotometer (Santa Clara, CA, USA) equipped with an integrating sphere and BaSO₄ as the reference. The bandgap of the materials was calculated using the Kubelka–Munk method. The photoluminescence spectra of the synthesized materials were measured using a Scínco FS-2 photoluminescence spectrophotometer (Instruments Solutions, Breukelen, The Netherlands) with an excitation wavelength of 299 nm. X-ray photoelectron spectroscopy (XPS) analysis was carried out using a Thermo K-Alpha spectrometer (Thermo Fisher Scientific, Waltham, MA, USA) equipped with a monochromatic Al Kα X-ray source (1486.6 eV). Samples were placed in a pre-analysis chamber and mounted on aluminum strips attached to stainless steel sample holders. The analysis area had a diameter of 400 µm. For survey spectra, a pass energy of 1 eV was used. High-resolution spectra were acquired by adjusting the pass energy to yield approximately 200 data points across each elemental peak, resulting in energy step sizes ranging from 0.005 to 0.02 eV, depending on the element and its corresponding energy region. The photoresponse capability, as well as the semiconductor and electrical properties, was evaluated through (photo)electrochemical characterization under conditions similar to those used in the photocatalytic hydrogen production tests. A suspension containing Nafion, ethanol, and the photocatalyst was prepared and deposited onto an FTO substrate using the drop-casting technique. The measurements were performed using an AUTOLAB PGSTAT302N potentiostat/galvanostat (Sciospec Scientific Instruments, Bennewitz, Germany) equipped with a FRA module, employing a conventional three-electrode cell adapted for the illumination of the working electrode (FTO/photocatalyst). An Ag/AgCl (3 M KCl) electrode and a graphite sheet served as the reference and counter electrodes, respectively. The electrolyte consisted of a 0.1 M Na₂SO₄ solution in a methanol–water mixture (1:1 v/v), and a 4 W LED lamp was used as the light source to illuminate the FTO/photocatalyst plate in a back-illumination configuration [56]. Mott–Schottky plots were recorded via potentiodynamic analysis at a frequency of 500 Hz.

3.6. Photocatalytic Hydrogen Evaluation

A glass reactor containing a methanol–water solution (1:1 v/v) was charged with 50 mg of catalyst, and the mixture was magnetically stirred under a nitrogen flow for 15 min to displace the dissolved oxygen. The solution was then irradiated with four 4 W LED lamps (λ = 450 nm). Hydrogen evolution was monitored using a Shimadzu G-08 gas chromatograph equipped with a thermal conductivity detector (TCD) and a Shincarbon packed column (Agilent Technologies, Santa Clara, CA, USA) (2 m length, 1 mm ID, 25 mm OD), with nitrogen serving as the carrier gas.

4. Conclusions

The CdS/TiO₂ composites were synthesized via impregnation and photodeposition methods and assessed for hydrogen production using methanol–water solutions under visible light (λ = 450 nm). TiO₂ modified with 20 wt% CdS exhibited a significant enhancement in photocatalytic hydrogen production, with increases by factors of 7.7 and 5.2 times for the impregnation (TiC20) and photodeposition method (TiC20FD), respectively, compared to CdS alone. The improved photocatalytic performance of the material synthesized by the impregnation method can be attributed to enhanced charge transfer, resulting from the better interaction between CdS and TiO₂. The impregnation of Cd²⁺ followed by sulfurization reduced charge transfer resistance, leading to a more efficient interaction between CdS and TiO₂, thereby improving charge separation. In contrast, when CdS is solely photodeposited onto TiO₂, the lack of sufficient interaction between both crystalline structures increases charge transfer resistance within the composite. This study underscores the crucial role of the synthesis method in optimizing the efficiency of materials for the photoreduction of methanol–water solutions in hydrogen production.